LIVE from SaaStr Annual 2024- Legends Stage A Stage - Wednesday

Ra.

Sa. Sa. Sa.

Sam.

Sa.

It.

Sa.

Ladies and gentlemen, please take your seats. Our show will begin in a few minutes. As a courtesy to our speakers and fellow attendees, we ask that you please silence all devices. Please welcome to the stage danny grant, CEO at jam dane connect, svp emerging technology and incubation at cloudflare, vincent vandermeulen, software engineer at figma, mario rodriguez, cpo at GitHub, and diego zach's vp design at ramp.

The comfy. Get a comfy one.

This is unbelievable. We have here an all star panel. In my mind, this is like if we had the Mark Andreessen's in the world in 1995 talking about the Internet. These four have built some of the most widely used AI features in production right now. And we're going to hear the real stories behind them. What we're going to do is the first half, each person is just going to tell the story of the biggest AI feature of their company and then we're going to discuss as a group the critical lessons learned. Y' all ready?

Yeah.

We're starting with Mario, who is the CPO at GitHub and personally helped build Copilot to where it is today. Copilot has 1.8 million paying users from 77,000 companies. It is the most used AI dev tool in production today. And you're going to hear the story.

Thank you. How are you doing today? There you go. Still filling out. So maybe more coffee. I'm on my second cup today already. Today I want to talk to you about Copilot. It's kind of interesting. Scott Chacon, which is one of the co founders of GitHub, recently posted. I think it was like two days ago a post and he said GitHub won back in 2008 because of two things. One is it was the right product at the right time and it was Git. It's kind of where everything was coming together, where open source was coming together and working in the decentralized way and none of the centralized version control systems could actually handle that. And the second thing that he listed off why GitHub won was because the founders had taste. And that taste actually talked about them being dev centric, them being developers and creating a tool for them that has been at the heart of GitHub since day one. It's this developer love this obsession about what do developers need to be successful. I have been doing dev tools for 20 plus years now. Actually I'll probably retire doing dev tools. I love doing that for a living and making tools for all of you, if you think about it, humanity already invented the fire and the wheel. So everything else we're going to be doing from here on is going to be powered by software. So I get a lot of energy about giving you all the best tools possibly. And that's kind of what happened with Copilot at the very beginning, if you think about it. The first time that I saw it was in 2021. This was before ChatGPT was a thing. And we were just playing with this model called codecs. And that Codecs model was a technology. It could do, like, really interesting things when it comes to natural language, but we didn't know what to do with it. And that's actually a challenge we face constantly in this technology world, which is, you got something cool, you have a tech demo. How do you turn that into a product? And what we ended up doing was kind of obsessing about that developer experience. We had introduced ghost text. We, we work really, really hard in making the latency be 100 milliseconds, 150 milliseconds, no matter where you are in the world. And then we ended up doing a lot of prompt engineering. So if you are in the audience and you're great on prompt engineering, I want to talk to you and hire you. But, like, it's because it's a real thing. Like, it takes a lot to actually do the pass the right context to these models and get the right answer. And that's how. That's how we are. And where we are today is by generating value not through a technology, but a product. And the story of Copilot is that it's one of being in the right place at the right time and also having that taste and really being a steward of those developers and the people using it. So what is the future then? And for us, the future looks a little bit past the developer in this case. And I was recently in a podcast where I was saying we need to be able to reach a billion people and make them software developers, and you're not going to be able to do that overall by being professional developers. So what I'm really excited about, kind of the next generation of Copilot is transforming into this product that your grandmother, your child, or you yourself, even if you're not a programmer, can actually interact and create something overall. And if we're able to do that, I think we're going to really accelerate humanity forward. So thanks for that.

Amazing.

Next is Diego Zaks. He's the VP of Design at Ramp, which is a financial ops platform. But probably all of you have heard of ramp because they're one of the fastest growing SaaS platforms in history. They reached 300 million of ARR in about three years. 25,000 businesses use ramp and they use AI to save those 25,000 businesses about a billion dollars and 10 million minutes. And incredible what they built.

Yeah. Well, thanks for having me and thank you all for coming. Who here is expensing this trip? Yeah, amazing. That's what we do. We make that super painless for everyone. We use a lot of AI for this. We build software that helps companies spend. All non payroll spend goes through us. So we serve the finance function, but we also serve the employees. Right, the people who are spending to do the job. So we see ourselves as a really a company that sits in the middle of that relationship between finance and the rest of the humans in the company. And it's easy to, when you're submitting an expense report, it's easy to forget that finance people are people too and they just want to help you do your job. So we make their lives a lot easier and hopefully make all of your lives easier as well. We, we've been using AI for from the beginning and it's really been at the core. But I do want, you know, in all of this AI stuff we talk a lot about the tech, the AI itself. And I think for us at least, AI is not the product, the product is the product. And AI is one of the ways in which we make people's lives easier. So the main thing that I try to focus our practice around AI around is does it feel like magic? And magic could be achieved through AI, like some really complex thing. But it's also sending an email at the right time to someone. There's putting the button in the right place and having like a really tight piece of copy that just makes something super easy, super easy to understand, painless for someone. So if it feels like magic, it's AI in my book. So in order to get to that first we root in real user problems. We make that technology, the experience disappear. We ask not how can we make the experience better or like slightly faster is like how do we make it go away? That's really where AI shines, where hey. And I'll show a couple of examples of that. Then finally, how do you operate at the right altitude between hey, the AI is just being helpful, providing a suggestion, saying hey, you might be missing some information to fully doing something. For me, there's a huge range between you might want to book a flight to go to Saster to. I just show up at the airport and it's taken care of for me. That last bit that I was talking about, that's where we're kind of exploring at the edges of what is possible with technology. So in order to do that, we hire just the best engineering talent, people who are builders, people who are curious. And we make it low risk for them to just play with AI and try stuff. So we have a range of things. So if you're expensing dinner, you just spend on a ramp card, we look at all the data of the merchant, the network, the company you're spending, we generate some memos, possible reasons for that expense. You just reply with number two and everything's done. That's it. So AI is not here, it's just doing it in the background. And then you go from the more obvious things to the much more magical things. So that expense was categorized by the accounting team. You upload an invoice, they're super hard to read. We transcribed the whole thing to how much did we spend on marketing? And instead of just very confidently spitting out a number, we produce and manipulate the ramp UI so that you can actually see the data and get to that answer yourself. And then it's kind of like a first draft of a report or, or a spreadsheet. And then we also have those amazing engineers who, when paired with a designer, get to explore a lot of experimental stuff. So if you've ever had to do a thing 50 times on your computer, by the time you've done the third one, third round, you're like, come on, why can't you just figure out that I'm going to do this another 40 times? So that's what we built. Once you do something on ramp a couple times, we essentially predict the next action and instead of clicking around, we just make it a keyboard command and just hit Enter, enter, enter, enter, enter. And just get through the list. And here's where, like, AI is not always the solution. The problem here is not that it takes a lot of clicks to do 50 things, is that we're making you do 50 things. So what if we don't even. That's the actual solution. Don't make me do 50 things at once instead of, you know, trying to AI suggest it. But so big range, let people play and just pair them with designers to figure out how to bring that tech into a real user problem. And yeah, explore.

Incredible. Next is Dane Connect. He is the SVP of emerging technologies and incubation at Cloudflare. Cloudflare is one of the largest computing networks. They it's a $30 billion company, a third of the Fortune 500 uses them. It's quickly become one of the leading AI clouds, where millions of developers are deploying AI apps. And Dane's only mistake in his whole career was 10 years ago, hiring me.

Thanks Danny and thanks Sasta for having me. So yeah, I mean a lot of you guys know what Cloudflare for. Sitting in front of a large portion of the Internet and protecting it from DDoS attacks and helping make it faster. About 10 years ago we started actually opening it up so that we wanted to be able to make it so that everybody could build applications the way Cloudflare builds them, where we don't have to worry about scale, we don't have to worry about picking a region for where it's going to run and really making it a cloud, where not just we can build our internal applications, but everyone here can build those applications. And in the past year, you know, we continue to add primitives for how you can build these full stack applications. We've deployed things like databases, storage and then most recently we added GPU storage so you can start to build AI enabled applications. So whatever you write in GitHub, where GitHub makes GitHub copilot, makes it easy for you to write that software and focus on building your business logic and get your applications out faster. We want to be the place for you to deploy those applications. And we hide the infrastructure, the scaling challenges that you might have so if your app goes viral, you won't have to worry about it falling over when all of those users are finally there to see what great things you've built. But when we started building, deploying, adding the AI primitive, we were kind of trying to figure out where's the best place that people really want to run AI. And the answer was a lot of actual places based on the circumstances. So you know, if you have a really powerful iPhone, the best place is a device, but a lot of people don't have those powerful iPhones. A lot of places where you want to run this stuff, it's not going to be there. So the next best place we figured was going to be the edge. And so that's why we made the decision to deploy the GPUs to the edge and make it the built into the workers platform, our serverless platform where you can quickly call any of the open source models, deploy your models and run that workers inference right there at the edge or Fall back and use our AI gateway when you need a route back to one of the larger LLMs like OpenAI to ask some of the larger questions. So your app can kind of handle all different types of the AI workloads that you possibly need. So yeah, in the past year we've retrofitted our 300 cities, 500, over 500 data centers with GPUs. Over 170 of them have them today. And pushing out more, bigger, faster inference every day. And really excited to see what you guys build. And really kind of what you said at Ramp really kind of resonated was AI is become really powerful when you're building into your application, not as a chatter face, but really as something that disappears and becomes part of it. And want to help you guys do that.

Amazing.

And lastly, we have Vincent van der Meulen, who is leading AI efforts at figma.

Hey, everyone. It's going to be hard to close out this amazing lineup, but I'm going to try. So my time at figma actually started outside of figma as a fan. So I was at the design tool startup called Diagram, and we were making these AI plugins for figma and one day Dylan called and he was like, do you want to join? And for me, it was kind of like being invited to Santa's workshop. Of course, like, I want to be where the magic happens. So Diagram decided to join figma as of last year, and now I get to build these AI features at Figma itself, which is truly a dream come true. So last year we set about creating the first AI features for figma. And where do you even start? Well, one of our first projects that we did as a company in the AI space was Jambot. And this actually came out of a hackathon. And the idea was that you can go to ChatGPT and you can brainstorm and you can riff. But how do we make it more fun? Can we make a widget for our online whiteboarding tool Figjam, for that lets you interact with ChatGPT and other APIs in a much more expressive, visual, collaborative way. So we built this widget that makes it a ton of fun to interact with ChatGPT, and that was our first foray into AI, and it really made us want more and made us very excited about what we could do in this space. So we started there and then for the past year, we've been wondering, we've been very focused on building our first AI features and creating this bundle for more professional designers. So when we decided to set out when we made this bundle, we really wanted to focus on, we really wanted to focus on complementing designers and not replacing them. So we asked ourselves, what are features that designers might want? So one of the features that we focused on was AI search. You know, as designers, we lose designs all the time. Can we help you find them faster using AI? Another one was make prototype. Creating prototypes takes a lot of work. What if we can get you 80% of the way there using AI? And then of course, we've all had the pain of naming our own layers. What if AI would just do that for us? So this is our AI bundle and there's many more features and, and we're now excited to iterate on these. And when we think about our AI features, we really think about a few key principles that we always try to adhere to, like Django and like Ramp. We really try to root this in existing user behaviors. So we try to ask ourselves, what problems are users experiencing and is this even a problem that AI needs to solve? And then from there we try to build these prototypes. And the key there is really frequent iteration. So we have an internal figma build that we ship experiments to constantly. We also do internal research, we do external research. Because once you have a prototype, it still takes a ton of time and a ton of like living with the prototype to get it right. And then of course, AI is notorious for not having the right level of quality. So how do we actually evaluate and make sure that these AI features have enough have the right quality to begin with? So a big focus is evals. So we set, for every feature that we build, we set up these little tests where we have the expected perfect outputs, and we then compare that expected perfect output to the AI outputs. And we do that for every change, and only when the expected output is close enough. Only when the AI output is close enough to the expected output, that's when we actually put these features into the wild. And that's how we kind of combat these AI features, sometimes being unreliable. And then lastly, we really think that AI is a space that we should all be involved in. So it's a very like multidisciplinary effort. So I myself am a product engineer, but we also have designers, we have ML engineers, we have researchers, a ton of infra people, and together we 1 think that leads to better ideas, but we also think it truly helps making these features great and keeping them up as they scale. So I'm excited to chat about these topics today.

So you've now met the incredible panelists by way of introduction. I'm Danny. I'm the CEO at Jam. We help 135,000 people fix bugs faster. Some of the largest companies in the world now use jam. Probably if you use the Internet today, you've used a website built with jam. We're in the business of helping people build and ship grade products. And here we have people who have built the most widely used AI features in the world. I'm so honored to get to moderate this panel. And the way that we came up with the questions is we asked every single one of them. What questions are you dying to know from the other panelists? What are the tactical things that GitHub wants to learn from Figma and vice versa? And so those are the questions we have for you now. Are you ready? So AI is changing really fast. Also, all of you work at companies that do planning. And so how do you do product roadmap to be able to adapt, do things on the fly as things are possible, but also can have a roadmap meeting and be able to forecast what you're going to work on this year, this quarter.

Who wants to go first? I'll do it.

Okay.

So I work in products, so this is something that I live and breathe every day. And what we try to do at GitHub is this concept of, of strategic roadmaps. And in the strategic roadmaps, you don't have, what are we doing next quarter, what are we doing next month? Or anything like that. You have more. What are these bets that you're going to be taking, and what is the question you're trying to answer on those bets that informs the next stepping stone of it? So it has this cone of what I would say, confidence of where you're going that is always rooted then on the strategy. So what we try to do is say, okay, this quarter, it's going to look like this. And we're trying to learn X, Y and Z. And based on what we learned from this, we're going to have this hypothesis. And the next quarter might look like X and Y and Z. And we try to plan out in a strategic way more or less a year, knowing that anything that we actually say four quarters from now is completely untrue. And if we actually are doing really good in sticking to that, then we played the lotto probably very significant amount of time. So that's more or less the way that we give ourselves. And even with our customers, we have a lot of enterprise customers as an example, and they want predictability. But in AI, you cannot have predictability LLMs are not predictable even on that. So on that end, what we try to do is making sure that we are achieving the right value in the product through a strategy and. And then learning constantly about it. And then we revisit that strategic roadmap in what we call npr. So monthly product reviews, and sometimes it's every six weeks. It doesn't have to be every four weeks overall. And we try to ask the same question, okay, what does the next quarter look like and what did we learn and how can we affect it? And that's how we try to stay more or less as nimble as a startup, let's say. But I still have the scale that we have today.

So Cloudflare does sort of similar where the farther you go out, the less fidelity you have. We divide it into three different groups, where we have a kind of a core product organization that really is focusing on what do we need to ship to customers next quarter, the quarter after that, what does sales need to drive revenue, to reduce churn, to make the customers happier. And we kind of call that kind of incremental innovation. Then we have another group my team runs is kind of the second horizon, where we kind of really focus on growing TAM and really trying to do things that are kind of 12, 18 months out. We're focusing on what are those kind of next big strategic pillars that we want to start experimenting with, investigating and starting to gain a better understanding of. And then we even have a third horizon where the research team which really focuses on not even with the idea of how you productize something, but really just kind of more of the fundamental technology of, you know, for us, how you build a better Internet. And, you know, a lot of that work over the past few years has been in the AI space, in which that will all kind of help inform the roadmap of the other two horizons as we go forward. So I think it's, you want to know where you're going long term, but you also want to remain agile. And, you know, I think every four or six weeks, really, if you're not reviewing the roadmap and really asking yourself, is the next thing that we're building the right thing, you're probably waiting too long.

I think one of the things where. And I think you guys are coming from the product world, Vincent and I coming from the design world, I think designers are very comfortable doing 99 things, finding 99 ways that something does not work before something makes sense. So at least for us, design being such a core part of ramp, we're very comfortable not really knowing where we're going to end up 18 months from now. We do know what we want to do in the next three months. That's just execution mode. We don't actually know exactly how that's going to work out, but we got that quarter locked down. But understanding that, that we're making two assumptions. One, it's okay not to know exactly what shape that's going to take. And that's part of the core underlying assumption that we're going to figure it out. And that same kind of optimism that's backed by people who work really hard and are actually super smart. We apply the same thing to AI. It's like, well, this thing doesn't really work today, but in two weeks there's another model, and it might work then, and might work then, might work then. So even if the latency is like 800 milliseconds now, we're probably going to be able to get it to 100 within the next four months. So it's really about focusing not so much on the things that we cannot do, but on the things that we want to do and being highly optimistic about what we can achieve within that year, rather than trying to figure it out, decide today with the limitations that we have today, which in the world of AI, that just leads to complete failure within weeks.

Yeah, I think it's the same for us. One thing we like to do is kind of at least somewhat like, skate towards where the puck is headed. So we try to, and this is hard, but we try to somewhat predict, like, okay, what do these AI models look like a few years from now? Right. Like, it doesn't take a genius right now to figure out that in a few years these LLMs are, are going to understand video. So how might that affect us, our roadmap, and what we want to do? But we also try to stay very flexible. So, like at Figma, we have a lot of hackathons. And one of our big hackathons is called Maker Week. And when Maker Week happens, we go into that week of hacking with the full expectation that this might completely change our roadmap. And in fact, right now, like the big project I'm working on, that I expect to now work on for the next year, actually came out of a hackathon like four weeks ago or something. So just to show you, like, you don't have to have these plans set in stone. They can change and give your employees the opportunity to actually come forward with these great AI ideas, I think one.

Thing I kind of heard kind of cross is also giving people and teams room to experiment. And I think it's even more important with the speed of the AI technologies moving today, where you get people on your teams a chance to actually not write long docs and sit in ivory tower and say this is what's going to happen, but actually go and say, how can I use these and touch them, use them, feel them. So making sure that your teams have those tools so that they actually can experiment and the time.

Dane, what does that look like in your team? How do you give people time to experiment tactically day to day?

Well, so, you know, I think early on we really start with very small teams where we give two or three people that are passionate about an idea just really the chance to just go and build something and prove something out. I love demos, so I love watching a team of a couple engineers and product people and designers kind of go off and come back in a few weeks with the greatest possible end to end demo that they can do. So like actually proving out things from the start to the finish and seeing how fast they can do that and seeing what they learn and then really investing behind that. So you know, which kind of goes back to the experimentation. Like you need to invest and create a place where it's okay to kind of fail because the failure might be that it might not be a failure of the idea might be a failure. Oh, the models just aren't there yet and you need to wait for the technology to catch up, you know, and you know, speed things are going. It will.

Can I riff on that a second please? So I think Vincent was talking about evaluation and at GitHub we invest in this thing we call coffee, which is copilot offline evaluation. And that's one area that I have found to be a place that you have to over invest in AI, if you will, if you need to experiment. Because you could experiment all day in AI and not accomplish anything without being able to understand how evals are going. So for us, the way that we kind of translate and we give the ability and the autonomy to the teams is saying anytime you experiment, you're doing offline evals and move quickly through that offline evolve place. Because even if you get an offline 95, by the way, usually that means the product, when it gets to market, it's a really bad product, at least in AI world, in my opinion. So you're going to have to then do online evolve to try to figure it out. Is it going to fail 50% of the time or 30% of the time or 20% of the time in the user. So that's the way that we kind of encourage a little bit of that. And we also then have a GitHub Next team that looks at Verizon 3 items or Horizon 3 items, like Dane was saying as well. So that's kind of a little bit.

Of how we do it.

What does that look like in practice?

The GitHub Next or coffee?

The eval? Do you have a team that's doing a test suite or like, how do you implement this in an organization?

Yeah, so it's not that I would say it's very analytical is what I'm trying to say is like you have this thing, you end up figuring out what are the test cases that you need to go and run and you need to pick the right data. So you need people that are in applying science that really understand that. You need product and engineering to understand, okay, how we're going to actually then create the evaluation to be able to score on that. So it's kind of very similar to benchmarking, if you want to think about. I don't like benchmarks out there, by the way, because you could game every single one of them, in my opinion. But what I do like about benchmarking is that it gives you a view into a set of scenarios that then you could then figure out are you getting better on those scenarios or not? So I think about offline evals mainly as benchmarks on are you going in the right direction with your hypothesis or not? And you just run a lot of them. Just recently I was reading a post by one of the teams and what they were doing is a lot of these models right now can actually do function calling. And 4.0 is really, really good at being able to say, okay, I'm going to call this function instead of this function. But when you have 30,000 functions that you could call, and that's a lot, we don't have that many. But imagine that world, then how do you make.

Then the model?

Pick the right one becomes a really interesting challenge. And we were trying with can we create agents in between with domain expertise. And then 4.0 can just go and pick the right agent and then this agent figures out the functions on it. So you end up running that through evals and try to figure it out, hey, is the hypothesis or not? And you get back, you know, an answer usually within two to three weeks on some of those things, and then you just move forward from there.

What's so interesting about what Mario's talking about is testing in software is kind of a solved problem. We sort of know how to do it. We can come up with a list of edge cases we can run through tests, we can even automate our tests. But with AI, it's non deterministic software. It's really hard to know is, is this ready to go out to users? Does this meet our product quality bar? I'm curious if any of you have any other good systems that you, you recommend or use internally for when AI is good enough to ship?

Well, so we, yeah, I wouldn't recommend this per se because it's a lot of work, but we build a lot of our own eval systems and I think it's important when you do evals that you do your evals in a way that matches your product. So figma is a highly visual tool. So it wouldn't come to anyone's surprise that we do visual evals and that we also do our evals inside of figma. So for instance, for our search feature, we would have the AI search model just spit out all the search results on the figma Infinite canvas. And then we could go in with a Figma plugin to actually rate those search results as good or bad. And we would do that every time we'd update our search model. So it's a really fun visual figma like way to kind of inspect the results. And I. Yeah, and that's the benefit of building your own evals, eval tooling. You can really make it fit your use case and, you know, optimize what's important, which in our case is being able to zoom in on the design and being able to see, hey, is this actually a good result or does this just look like a good result at the service level?

One thing that. One thing that all of us in this room are trying to figure out for our own products is how important is it to be first to market in AI versus sit back, watch the market, see, see what works, let everyone else do R and D and then build it as well. And curious for the four of you, how do you think about that in your own organization?

I mean, I strongly believe that you have to put yourself out of a job every 18 months or someone else will do it for you. So that is the philosophy at reb. It's all about velocity. It's about getting there first and second and third and fourth and fifth. That's really the key. In order to do that, we set up basically two key beliefs, which is all we do is betting as long as we're right 52% of the time, we're going to win. So make a lot of bets very, very quickly. And the second is make it really cheap and easy to correct or when you're, whenever you're wrong, it's really easy to recover from and just do the next thing. So it's really everything around Ramp is geared around velocity of our ability to just make changes to our product instantly, every day, multiple times a day. And when we see something in the market that works, that someone got right, we have the muscle to just like literally go do it. And two weeks later we have something that kind of works or doesn't work at all. But then two weeks later it starts to work.

Cloudflare is very similar. Actually spoke with your CTO a little while ago about this. You know, we focus on velocity and shots on goal. You know, you want to get as many things out there as possible, get early user feedback, know that you can continue to improve it over time. But until you hear what the customers have to say, say in a lot of cases you're just guessing what the reaction is going to be. And you know, especially now that, and I think that velocity is only going to increase with tools like Copilot and all the developer tools that are being AI enabled today.

The alternative to what you're describing is what I see happen by default at a lot of companies or most companies, which is as people progress in their careers, they want to start making only good guesses, only the right guess. I want to be right most of the time, I want to be right all of the time. And then what that creates is like it just takes you six months to have some confidence or more confidence on a guess, but you might still be wrong at the end of it. So it's this like sort of self fulfilling prophecy of if you're not just fearless and wrong all the time, you're going to still be wrong, it's just going to be super slow and probably someone's going to take your job. So it's.

Yeah, you also said something about, you know, specifically about AI features. And so I think that's actually from a product person's perspective, I think that's kind of a mistake. I think yes, marketing can slap AI on any of our stuff and hopefully, you know, get more attention and ride ways. But from a product perspective, the product teams focus. You know, I like to encourage them to what customer problem are we solving? And AI happens to be a great tool in our tool chest these days. That it wasn't there a few years ago. We should use that as much as possible to help solve those customer problems faster, more efficiently, better, but not necessarily think about it as how can I ship AI features faster? How can I solve customers problems faster?

I mean, I think we similarly like want to move fast. But at the same time, like, I do think quality is very important, right? Like figma for instance, is a tool you use on a daily basis. Like if we start shipping AI features that don't meet your quality bar, very frustrating experience. So what I do like about what we do is that we have this plugin ecosystem, right? So even if sometimes we take our time to really get a feature right, we can encourage our plugin creators to fill in the gaps, right? So we have something like 200 AI figma plugins right now. And I of course myself used to make them. So I'm sure GitHub might have something similar with their like extension ecosystem that, that could be a powerful way to kind of like strike the balance of delivering quality features while also keeping that speed and like giving your users like, AI value on day one.

So a lot of us here are building out applied AI teams in our companies. You all have done that very successfully. What's the recipe for a great internal applied AI team? Who should be on this team? What should be the makeup of this team? Who's fit to lead this team?

I'll start this one then maybe what I would say is there's two components in my opinion, creating an AI product. One of them is the applied science component. And look, there's not that many PhD people out there in the world. The universities do a very good job of constraining that inflow. So you definitely need PhDs in my opinion, in many of these fields. They're really involved overall. But one of the cool things is like having these LLMs like OpenAI is innovating. You get anthropic innovated, you get Mistral innovated, so you don't need access. They actually democratize a little bit getting like having to invest in very deep applied science and very deep in MLOps, in my opinion. So like that's one of the benefits. So I think you do need an applied science team. You need them to actually have a significant amount of understandings of how LLMs and neural networks sometimes work. You need people that are very good in data and can act, you know, create these evaluation suites and all of those items. But you also need, and you started to see this more AI engineers, which is you take A developer today and now they have to inject intelligence into an application. In all of our applications now have intelligence, which means that we need to then move where the traditional developer is and then allow them to innovate on top of this technology overall and interact with a bigger ecosystem that they were not used to before. So for me the right thing is the not to say you need a bunch of applied science. I think you also need to grow the company to be AI engineers plus applied science. And that's what at the end might create the right product overall instead of just investing in the deep tech.

So for example, what is the copilot team makeup sort of look like?

We probably don't talk too much out there about that. It's not like a super secret sauce, but it's kind of interesting. What I would say is we try to nasty centralize it. There shouldn't be just one copilot team, there's a copilot intelligence platform team. And that's true. But I would say the repos team is a CoPilot team. The PR team is a copilot team. The issues team is a copilot team. So the, you know, people that we have in VS Code and ide, they're in the copilot team. So I think that's the way that I want that I would phrase it both internally and externally. And so all the companies kind of center now on that copilot.

And when you talk about building AI engineering as a skill across the entire organization, is there like a reskilling program happening internally? Are you hiring for that? How does that happen?

Yeah, maybe you all can answer that and I'll answer it last. I'm interested in. Actually that was not in the questions we had to.

Sorry.

So for us it's like we look at like enthusiasm a lot, right? So like we have some dedicated ML engineers who have been doing ML for a long time. But a lot of us are just engineers from all over the company who are so excited about AI and that then came into our org to build these features. So we have some, I guess, reskilling programs or whatever. We have some. Sometimes we have certain days where whatever research will share out research and there will be some education. But what we'll see is that a lot of these people kind of reskill automatically because they're so passionate about AI and are just constantly building side projects in their own time that like we don't even need to do the reskilling. These people just are bursting out of the seams of ideas and they're like toying around with Copilot and all these other tools in their free time and we just need to give them a place to shine and yeah, that's a lot of our AI engineers, to be honest.

So what, what's sort of the makeup of AI at figma? Is it team? Is it.

Yeah, we, we, we, we have a mix where I think we're, we're also starting to realize that like, okay, we do need like a, yeah, fundamental team of like, ML engineers and like people who really can like build our, build custom models. But then, yeah, we do have a lot of like a. Yeah, a lot of like AI product engineers like myself. We, yeah, we have designers who are like specifically focused on AI. And then, yeah, what I haven't heard a lot about is research, which we think is crucially important. So we have a lot of researchers who will basically focus on doing a lot of research with our experimental AI prototypes both inside and outside the company. So we can get a pulse of how our customers responding and what is the quality actually like. So it's truly a very multidisciplinary disciplinary effort at figma.

I think one of the key things to remember, at least for us, is AI can be customer facing, but at least in my experience, it's been mostly internal facing. That's where the biggest transformation has come. So we know what we need. We have this many support specialists, we have this many customers. How do we make each person more effective? How do we build AI tooling for them? Everything. So our underwriting models, our support system, our operation system. So we've, we've actually just have spent most of the things that actually hit production that are more experimental are internal to transform the way that we operate. So I think what we do is we, we have an applied AI team who are just constantly, all day, like plugged into whatever is happening on AI. They know about it like the day, the day before it goes out. And they are basically the inspiration for everyone else in the company to say, this part of my job really sucks. I would like it to go away. And then they can say, oh, try this model. And then we give that person the tools to actually just go do it, just figure it out. And then they try. So we really look for people who are builders who are curious and are going to figure it out on their own and whenever they get stuck, they're going to raise their hand and say, hey, I need, you know, I need help. We obviously have parts of the product that can't break. We move billions of dollars cannot mess with that. You don't want to experiment there. But we also have product that it's fine, if it breaks, we fix it right there. So it's really a balance of what can you break, what can't you break? And if it's. It's really risky to make a decision, you have to think about it more. You have to get the experts, the people who actually know really what they're doing. But everyone else, they're going to get pretty far on their own if you just give them the space and just tell them to go do the thing. A lot of times you just have to tell people like, hey, you can do something about this.

So I've talked mainly about workers AI and how we kind of give everyone else the ability to deploy work. AI enabled applications. The cloudfront we kind of actually think about in three different ways internally. One is operational AI. What tools are we providing to our employees so they can do their jobs more efficiently, they can build better processes, deliver more value to themselves and to the customers from an internal perspective. And recently we actually created an operational AI team that other teams can kind of kind of like a McKinsey consulting group inside for people wanting to go solve those problems like this just takes me too long to do this every day. How can you make it quicker and faster and kind of using a team that has some best practices to help teach that team how to build those applications. And then our product teams for the products that we actually ship to our customers. We've been using AI ML for years. Every single request that goes through us goes through a security model that decides is the request good or bad. And so there's that AI kind of proficiency is built into every team from our security products to the products that show in our analytics. And then obviously the third one is one we've talked about which is offering it as a developer tool for everyone else.

I have one last question, lightning round before we open it up to you all, which is imagining your products five years from now, what do you think it will look like? Is it the same product with AI features or do you think it's radically transformed with AI?

GitHub looks like TikTok. No, I'm kidding, that's hopefully never get there. What I would say is it looks more AI native. One of the beauties of what we're living through right now is that these LLMs can actually understand natural language in a way that none of the other tools before could. So I think if we really lean in into that and go and extract the maximum value of natural language to something that is innate to us as human, then the product looks completely different. And you can hold me to that. Five years from now, GitHub continues to look the same way as it is today. I failed you all.

I mean, for us, we actually measure engagement time and we want that to go down. So hopefully five years from now, nobody's using Ramp, it just works and it's just doing everything in the background. So that's where I think it's separating that experience from the people who actually use it to do their job and the people who have to do their jobs instead of it. So we want to make it go away for all of us and then have finance teams that have 1,000 accountants, a thousand AP clerks, a thousand like, per person. That would be truly the magic there.

Yeah, I mean, I kind of similar, I kind of hope AI kind of disappears as a thing that we talk about in the forefront and go back to the business value that we're creating. And I think in five years, I think the models are going to be exponentially better. Hopefully there's something even beyond LLMs that are more efficient and can better reason and can, you know, really just. I can't even imagine what they'll be able to do as far as enabling us to just do our jobs and live our lives better. Yeah.

I mean, five years from now, nobody of course knows, but I do think some level of role collapse is going to happen. Right. Just super excited. Exciting. So designers will be able to create software, software engineers will be able to create designs, and all of these roles are going to start blending together more.

But I think what we believe at.

FIGMA is that like, regardless of what happens there, like some fundamental, like some truths are fundamental and that's that this AI design tool of the future will probably be visual. You probably still want to collaborate with your co workers and we, yeah, we believe that regardless of what happens, yeah, we can kind of like invest in those principles and be all right and yeah, create something nice.

We've got here the product leaders of AI features at GitHub, Ramp, Cloudflare and Figma. We have time for two questions. Who's got the spicy questions of the group?

Hello.

Yeah, so I have a question for the Figma team. So what is sort of like the, I would say the line between which you decide, like, are we going to.

Push a AI feature or not?

Because like figma in the recent release announced a lot of AI features, but some of them didn't work. And then you have to roll it back. So what's that line and how do you judge that line when you are launching applications in production? And the same thing is for GitHub as well. Like you know, a lot of features are launched and announced but then rolled back.

Ooh, that's a great question. Yeah. What's that line? Well we try to kind of like anticipate where that line, where that line is. Right. So we have evals internally to see if the quality is good enough and then we do a lot of user research to try to understand like is this like something that our customers want and like is it helping them? Do they feel like this is like the right direction? But yeah, sometimes we will get it wrong and at that point I think the best thing we can do is iterate quickly and follow up with better AI features. I don't know if anybody knows where the line is but we try to anticipate where the line is and then whenever we get it wrong, be quick and just be in constant communication with the designers were actually making this tool for because if they're not happy then like what is this about? Like these AI features don't really matter if designers don't feel more empowered and if non designers don't feel like they can contribute to the design process. So yeah, everyone is still feeling it out but we're committed to getting it right there.

Yeah, I mean I don't remember of us rolling back a feature but definitely there is a line for us and usually that has to do with responsible AI for us and that's why we don't try to cross ever overall.

But other than that we try to.

Experiment in the open. If it doesn't work out, you know, we might go limited beta or something like that but the line is usually responsible AI for us.

Also a follow up question if you.

Don'T mind, we've got one last question so maybe meet us later.

Yes, thank you.

In the long term vision, how do you think about pricing and if you.

Have pressure right now about pricing from.

The business, about pricing per se or other ways.

I think the way we think about it is the trade off between how much do you value your time like you yourself. So if we can save you 30% of your time, it's worth something and if we go from 30 to 40, 30 to 50, it's worth something even more. So I think we currently don't have really any pressure. We're much, much earlier stage than GitHub, Microsoft or Cloudflare and Figma so we don't have that pressure today, but we are thinking about it as if we are able to save you 2, 3 headcount. How much does that actually create for you? How much of that is fair for us to take to actually invest more and then save you that fourth headcount?

For us, it's going to be gross margin related at the end is how we determine price and then market share, how aggressive we want to be on the combination of those two things.

We're out of time. Okay, you all, this has been amazing. Thank you so much and thank all of you.

Sa.

Ladies and gentlemen, please take your seats. Our show is about to begin.

Sa.

So it's been busy the last few weeks. Yeah, we wanted to make sure we.

Went public before this event. So.

Saster stunned people when it started because we didn't know there were all these CEOs that wanted to share. It was truly organic in a way that I don't think would happen today. Get those guys to share their nuggets.

Of information that they've spent 10 years learning is phenomenal.

There's a lot of young founders that are sitting in the audience.

It'll be really cool to look back.

And be like, hey, we were a part of that.

And this is when it all started.

Saastr isn't a conference. It's a community.

Saastr, to me, is home.

It feels like a reunion.

When you come to saastr, just walk.

The floor and meet every CEO and swap notes.

Saster has always been known for its content.

You could argue that this conference has.

Helped SaaS be what it is today.

I know any content that I'm engaging with from SAASTR is going to be worth my time. You will get out of Saastr what you put in.

It's amazing magic that can happen.

The impact that Saastra has had on.

Me is just to think much, much bigger.

There are very few communities that bring this group of individuals together.

You never know who you're gonna meet.

And what that might yield. That sort of connection is why I.

Think we come to these conferences today.

It's all a bit of an experiment. It's really about us helping each other. What do we want to do with this now that we've built this together sort of accidentally and organically?

Please welcome to the stage Jason Lemkin, founder and CEO at saster.

Thank you, everybody, for coming. How many folks have been to a Saster event before? It's a lot.

Okay.

Half.

It's good. You know, I'm going to talk a little Bit about where I think SAS Cloud, where we are today. But I really think we've split into a world of folks going for it. And folks may be phoning in a little bit. Quiet, quitting, working 10, 15 hours from home. So my biggest takeaway, when I got here yesterday, I sat out front and just talk to 100, 200 people, founders and executives coming in. And these were the Pirates and Romantics. These are the folks in this building, on this 48 campus that are going for it. So I think get to know each other, be there for each other because I think the founders and executives here are the ones going for it, which I think is great. And I think we've got to find a way to share our tips and tricks together because the world is bifurcated into the Pirates and the Romanics and the rest. And that's what I want to talk about. Where the bar is today before we get there. The next session after this. I'm doing a deep dive with Andrew Bielecki who is founder and CEO of Klaviyo. I would encourage all of you to stay. I only know him a little bit. We've done a SaaS tradition before. But he's one of my top heroes in SaaS. Klaviyo was the first to IPO after almost two years. The IPO market shut down in December of 2020 with HashiCorp. There was no IPO in SAS or cloud. SAS in 2022 and Klaviyo was the first after almost two years in September 2023. Now they're coming up on a billion in revenue, growing over 30% profitable. We'll talk through that. But they did, they did it the interesting way, was bootstrapped for the first few years, raised a bunch of cash, but then immediately got profitable again. It's a great story and there's so many things. So I think I picked two of my hero CEOs to do this year. Andrew from Klaviyo and Renee tomorrow from Bill. But come to these ones because I think they're, I mean anyone that's gone public, it's like the NBA, right? Anyone in the NBA is pretty good. But I picked a few of my favorites. So stay. It'll be immediately after this session and I'm going to talk about something I made up which may not stick after today, but hyper functional SaaS, which is what I think we're facing today, the bar has gone up. The customers are expecting more. Some of it's AI, some of its competition, some of it's budget, budgets in general, but it's my thought before we get there. Well, maybe we hit that one. Whoops.

Sorry.

Okay, the confidence screen is off. You guys. Can you fix this? So good job. Just a quick guide to today, tomorrow, and then I'll get into it. I think actually on this stage we've had a. We've had a tradition the last four years since we went to this campus of having some pretty, pretty fun folks for the big party. We had, we had cash, Cash. We had cheat codes. We had Duke Dumont, my personal favorite from Coachella. And then we have Titleau from Coachella. This year is going to be pretty bopping. So stay here. That will start here around I don't know when, whenever it says five or six. We'll start a party here then. And then for folks that have signed up, we're having a Croceo poker night right at the other end of the casino. If you didn't sign up and you are a CEO in Crow one again, there'll be a little queue line. You'll eventually get in, but. But we have, we are a little capacity constraint and then we publish it on SAS or on the website, on the blog. There's probably 20 informal parties, a lot of them in San Mateo that if you're not tired, you can walk to. But check that out. Go, go, go. As late as you want. Tonight, tomorrow, a couple of things just quickly. We have, we added a founder VC brunch from 11 to 1 in the founders lounge, which is just over in the other hangar if you haven't been in it, which is for founders only. So if you want, if you're ever thinking about raising money, if you want to meet VCs or all the VCs here that haven't done meetings, just get together 11 to 1. It's all about connecting founders and VCs. It should be a good concentration. Come for that if nothing else tomorrow. If you have any interest in fundraising or VCs, if you, if you want to get a deal done in the next 12 months, I know many have slowed down. Maybe we'll chat about that. If you're a CCO OR VP of CS, we have 200 folks coming to our CCO summit at the executive Briefing center down the causeway tomorrow. And we also about 200 CMOs and VP of marketing coming. If we tagged you on that, you got an invite. If you didn't get an invite and your badge literally says CMO or VP of Marketing, you can come and get in. Just show your tag and be a little bit patient. And then tomorrow we'll have a big wrap party here. We'll go until they kick us out. It's pretty fun. And if you're not tired of me, I will try to do like two hours of endless questions you can bring to everything that we've learned together for two and a half days here. And we'll go until they kick us.

Out of the venue.

Okay, so with that, a couple things that I've been thinking about. And I just got back to the Bay Area full time. I've been bouncing back and forth since that little bug went around in March 2020. Southern California, Northern California. I first went down there when Covid hit, mainly for family reasons. I came back up. The world was locked down. I've been bouncing around. Finally, I've only back for two and a half weeks in downtown Palo Alto full time. It's already better. It's already better. I have pumped into so many folks. I've seen. I've done three investments, I've seen 10 investors or founders. I haven't seen much in years. It's been great. I have had more interactions which sort of as an introvert, I thrive on more human interactions. Maybe in the last two weeks, in the last year in our community. And it has been great. But it's also different. It's also certainly if you've walked in parts of San Francisco, you'll see that it's different. The other day I was up there, I walked through the Embarcaderos. You know, every single store is empty in the Embarcaderos there it used to be, you know, white collar lawyers and some tech. Before March20, every store is empty. You go to Palo Alto, where we are, which is smaller. It's boppin'. I mean, it's good living. But if you squint, you know, it's still 19% vacant on the retail so small example. So it feels the same. If you look on Twitter, there's unicorn rounds left and right. Obviously AI spend is super exciting, but we haven't gone back. I say it's same but different. And a couple things to note. You know, unicorns are almost non existent. In 2023, they're up 50%. If you look at Carta, that's your other data. Unicorns are up at least 50%. But they're either big AI bets betting on massive growth, growing 10x or more in a year. People aren't some. There are some dumb AI investments out there. You may have seen one. But the truth is people are looking for epic growth. So if you see crazy valuations, there's either crazy growth under the hood, see OpenAI, or there's big promise of like literally top 1% growth. So that's driving unicorns. Or, you know, it's funny, you'll see someone, they raised at a billion last week, but they also raised at a billion in 2021. What happened? It's not just that it's a flat round, it's that they probably grew 10x or 5x since then. Okay, so the, the non hyped AI unicorns, they're often all at 100 million or above in revenue. Right? They're often 10 million at the peak of 2020. So the unicorns have come back. One is AI spend and some height. But also we just grew a lot. Like we up grew, grew a lot. And there have only been three. Going to my third point, there've only been three IPOs since 2021 Klaviyo. Andrew will be here right after this in 22 minutes. One stream, super interesting one that was bootstrapped to 100 million. SAS for CFO and compliance. They're at 500 million Klaviyo. IPO to 500 million. Rubrik IPO to just past 500 million. But Rubrik announced their earnings, I think yesterday. It's. I think it's on the next slide. I'll get a little bit wrong, but roughly 700 ish million in ARR growing 50% trading at 5 times ARR. 5 times ARR. 5 Times ARR. This is the best of the best. We had three IPOs and we'll have some good ones coming. We'll have the canvas and the stripes. We have the CBO of stripe was just hepar. We'll have some really good ones in databricks, but these ones were really, really good. I'm gonna ask Andrew why he IPO'd right now. And 5x for Rubrik. I mean, so the. It's the same but different.

Okay.

It's harder. And the fourth point, and I have a lot to say on these three points. The fourth one is one I'm struggling with. What I'm struggling with, which is anyone besides the folks here working in SaaS and Cloud, is anyone working. Does anyone want to work more than 10 to 15 hours a week? Does anyone want to come to the office? Not us. Half of us want to either have a hybrid office or at least be together in person part of the time. But does anyone else in tech, especially after all the change, the craziness of 2020, the boom of 2021, the beginning of hard times in 2022, the crash of 2023, and the weird rebound or whatever. We'll talk through that. But I think you've got to be relentless about bringing folks on your team that are real pirates and medics that want to do this startup journey. Because I just think the rest of tech doesn't want to work. And I'll show you, I know it sounds a little grouchy or whatever, but I'm going to show you a crazy data point in a second that we got from SAS Jones. It's crazy. It's crazy. And so I worry about this one. I worry not about us, but I worry about how we're going to and I worry not about founders. I think founders are as tenacious as ever, but I worry about some of the rest of our community. I don't have all the answers, so couple points I wanted to make. First of all, sorry, there's no downturn. There's no downturn. There's no downturn, but it's different. It's a new normal. So I know there's a lot on this chart, but I picked together, put together and left some of the points that it's not all a down if you think it's a downturn. If you're Debbie Downer and whoa. You know, for folks that are struggling on the right column, you can listen to their analyst calls and they'll be like, oh, there's dramatic macro impacts and our sales cycles have lengthened to 20 years and all these issues and it's all true in certain categories, especially B to B2B. I know this term doesn't quite exist, but folks selling directly to tech and others have been the most impacted. Folks selling to the end economy and B2B2C have been less impacted. But I just put together a list of some challenges just to understand what's going on because it's not a downturn overall, even if in segments you got hit by a sack of rocks over the last 24 months. So ones I call non tech folks selling mostly to end consumers or to SMBs like toast, growing 29% at a billion and a half in revenue Monday, which you know, Iran was here last year. Great session. They just crossed 34% growing a billion in ARR and super profitable. But 70% of Monday's customers, even though you might think it's like other folks and asanas and notions and other great products, 70% are outside of tech. That's the key. Shopify 21% growth at 8 billion. Shopify is accelerating. It is growing substantially. Faster than 18 months ago. Shopify exploded with COVID It went from 1 to 4 billion in one year. And then when the world reopened, it flat flattened. Now it's back 21% at 8 billion in ARR. Help me do the math. What is 20% times 8 billion? I think that's adding 1.6 billion of new revenue in one year. One point. Think about some of these numbers in absolute terms when you. When you're saying how hard it is out there to sell to merchants. Klaviyo. Andrew will be here. 900 million. Growing 35%. Growing 35% and very cash flow positive. Okay. This is marketing for e Commerce and B2B 2C. It's the HubSpot of E commerce. We'll talk with Andrew Y Samsara. EPIC1. EPIC1. We're going to do a Saster podcast with the CEO in a couple weeks. Just announced growing 36% at 1.26 billion. Selling to fleet management and truckers. And now other categories. It's expanded its verticals, but the trucks are still running, the cars are still there. We're still in 3% unemployment in the U.S. so Samsara is growing at an epic rate. That one's trading at 20x. I think so. So they're not all, but they're profitable. And listen, I don't need to put up Google Cloud, Azure or GitLab, but these are all folks tied directly or indirectly. The AI is booming. Google Cloud and Azure, huge rebound. Go talk to the folks at Google Cloud right there. They can give you more Saster bucks. Them and Oracle if you need more Saster bucks for swag, both of those. But those partners have it. But Google Cloud, massive acceleration, even Oracle, massive acceleration and GitLab, look at it. 30% growth at 730. Pretty good. And then the last category. And sorry to read from the slide, but I want to read what's behind it. Security and compliance are strong. I mean cybersecurity, goodness gracious. Wiz. I mean Wiz. Four years to 500 million in revenue. Right. And go off to a billion soon. Turning down apparently a $22 billion offer for Google. Yeah, because there's no downturn. It's not that it's necessarily super easy, but there's no downturn. So yeah, again, CrowdStrike up 34% at 4 billion. Even though crowdstrike, they had that little outage. I don't know if any of you noticed, my gym was down for a week, so still grew 35% at 4 billion. Rubrik, we talked about 35% at 120 million. The CEO just said, listen, there's no downturn for cybersecurity, but it's harder to get through procurement. It's more steps, there's more scrutiny. He did say that. But at the end of day the deal's closed. They're still closing. And then this one, stream compliance for CFOs a fun one. That bootstrapped to 100 million in ARR. 37% growth of 500 million. So these are not. I don't see any downturn. Anyone see a downturn on this column? Now if you look at what some of these folks are growing in 2021, it's not the same. Okay? Even the ones that are not seeing any downturn were growing even faster in 2021. So there are impacts like that, that insane weird times when we just bought everything in the world in 2021 because we're coming out of lockdown. That's gone forever. But there's no downturn here. But here's some leaders on the right. Salesforce has dropped for the first time ever to double digit growth. Now we all thought it had to happen at some point. They're coming up on 40 billion in revenue, but single digits. Mongo fell from high 30s to 13. So it's not that simple. In cloud Elastic, another cloud leader at 18 stock and Asana sort of a. Asana is a lot like a lot of us. You know they are selling to non tech but a lot of it's B2B2B growth has plummeted to 10%. So there is challenges. There is, there is maybe even a segmented downturn. But don't be thinking it's everywhere. And go find customers in the left column. Go look at these guys. Go figure out. We'll ask Andrew what's going on with their customers. Go find a few of those vendors. Go find a few of folks that, that are buying GitLab or these. You can't change who you are completely. But most of us have a heterogeneous customer base. So go focus on more and more of the folks that are buying more. I know it sounds obvious, but I wonder if we're doing enough of it if we're all complaining it's a downturn. Okay, so my first point is to say not that it's not real, but to challenge you that plenty of folks are growing at rates that make our jaws drop on the floor. But it's different. The multiples are Sometimes low and it's not necessarily easier. Where do I think SAS is going? SAS Cloud, AI. But where do I think it's going in 24 and 20 next five years? And you know, there's half, half of our content's AI. We have a whole AI stage in AI Summit. AI, AI. I'm going to talk, I'll talk a lot about AI, but I think it's more than that. I think AI a year ago, a year and a half ago in B2B was experiments. You know, David Sachs and I did a one on one last year. He's a little busy at. I think they had a summit, I think. But we did one last year and he said 80% of his 5 billion of craft ventures was going to AI. But I asked him what were the most exciting ones. He said it was mostly experiments, right. Early stuff. I don't think that's true today. It is crazy. If you look at anyone in the post sales space, in the CX or CS space and I'll show you a slide in a minute. Zendesk across AI has ripped through the entire category. Ripped through it. It is not an experiment. It is the core of the entire category. So it's quickly becoming table stakes. And what does that mean? I think what it simply means is the next point product has to be much better. Our customers are expecting it. Our customers are expecting that what you used to not be able to do, what you used to have to build your own analytics or build your own reporter, hit refresh or write, write a macro. The product has to do all of it. Your customers are going to expect every single one in your room to have AI in your product that automates things, automates things for him and in many cases pay nothing more for it. And yes, it's hard. And you're going to have to do the same thing with three engineers or 30 or 300 or 3,000. And you're still going to have to do all of this with the same budget and the same headcount because the customers have expected it and the public markets are expecting radical efficiency. We'll talk with Andrew about it. Klaviyo's coming up on a billion. They're incredibly efficient, huge operating margins everywhere. Everyone's gotten more efficient and you need both, you need growth and this. And then maybe the. We talked about automation everywhere. And then I think the most important thing is everyone now is probably going to have to go multi product earlier. We've learned this. We're not going to wait. And I see so many folks at 20 million, 50 million, 100 million, 200 million AR now stalling out. If you talk to a lot of growth VCs here, they'll tell you, God, I got this one that was hot in my portfolio. But they're at 100 million ARR growing 10% now or 200 million growing 15% and they were a rocket ship. Now some of it may be this so called downturn, but if you really pick at almost all these stories you're going to hear they just have one product, maybe they have some features, maybe they're trying to monetize a few other things, but they just stayed single product too long. And I think that plus competition plus AI means we have to go multi product even earlier. I know all the great YC advice is focus and I think it's true. But we also, we both have to focus on and go multi product earlier at the same time. The bar's up, it's hyper functional. Customers want more, they want more. And I think this is why a lot of folks that have point solutions are pulling their hair out, right? And you look at convergence, you look at folks like from Gong to Clary to Sales, off to Outreach to Calendly and on, and they are all very different products and categories. Four or five years ago, zoom info, now they're all, they all have the same stuff. They're converging, right? And there may be a lot of reasons for it. I'm not even sure I know the answer for that to that. You know, I asked the CEO of Gong@SAS Europa this year, I asked the new CRO at Salesloft why they thought of it. I don't know that they were 100% clear, but I think customers are just expecting hyper functionality for that same dollar, for that same nickel. It's not just that they want suites or they don't want point solutions. They're just expecting much, much more for that same dollar and we have to deliver it. So everyone's going to have to ship more code and work harder and build better software than 24 months ago or you're gonna grow 8% when you finally get to 100 million AR, it's not worth it. So it's life. The bar has gone up as it should every, every, every four to five years. So just a couple examples here. I say AI is a force name, but it's already becoming Table six. I just pulled up Zendesk's Intercoms and Marc Benioff's latest site. Like what's on their homepage AI first, okay, it's not an experiment. And not only that, it's table stakes. Like, you have to build if you want to be in the contact center now, if you want to be, if you want to be in this CX space or CS space, you can't be thinking about AI. You can't be working on it. It's already become table stakes. It is, it is absolutely mission critical and necessary. I'm on the board of a company called Gorgeous. I think the CTO is speaking today about how they added AI to their product. They've got 16,000 small businesses already. 20% of everything's automated through automation. AI already, already. It's table stakes. You could not roll out a product in this category without being AI and automation first. It's table stakes, right? And so I, I know folks get this, but most folks, many folks are still too slow here. They're still too slow here. They're still complaining they don't have enough engineers or it's too hard or it's too expensive. Cry me a river. Cry me a river. If even Zendesk is AI first. Okay, Salesforce, these are great companies, but I wouldn't say traditionally they're the most agile on planet Earth. They're getting it done, folks. Some of it may be more on the homepage than in the product, but they're getting it done. And so I think, and I know I said this, but this is my big challenge to everyone. I just think the reality is the bar to building a great product is just way, way up, way, way up. I mean, I think Klarna said that they're, I think they said yesterday, allegedly they're canceling Salesforce and Workday and they're going to do it all with AI. Okay. More power to them. More power to them. And maybe it's true, and maybe it's like 80% true, I don't know. But I'm sure it's directionally correct. But I think they're at the extreme. At the extreme. But customers are expecting to at least some of that. They're at least doing some of that. They're expecting AI to replace maybe 20% of their stack. And if you look at folks in the Context center, the Zendesk Intercoms, Gorgeouses, and they're, you know what they're expecting. I'll tell you. They want to replace 30 to 50% of their customer success workers with AI. It's already happening. It's already happening. They are laying off 20%, 30% of their contact center, not 100 like a klarna. But it's already happened. And so right or wrong, we could debate whether this is good. We could debate whether all AI is really answer your questions properly. I mean we all, we all have ivr scars. So what? This is really going to be so great. There's a lot to learn. We have the CTO of 5, 9, we'll have the CTO of gorgeous. But forget the reality. Customers are expecting this. You better deliver it. You better deliver a product where 20 to 30% of everything can now be automated through AI. You better deliver a product where if they, if they expect it, 20 to 30% of the workforce can be eliminated with AI. Right. I just did an investment in a company called Mango Mint, which is an SMB solution for salons and spas. Okay. They're at eight figures in ARR, growing over 100%. And why, like why is this old small category exploding? Well, it's cool software, but it turns out at these like salons and spas, they can't get anyone to man the front office anymore and they can't get anyone to follow up on the appointments. They just can't find any humans. So even the SMBs need AI and automation to do this and the customers are just demanding it. Once it's there, you can't compete. If your product doesn't completely automate the front and the back office, you can't compete. So now they've automated payroll and they've automated all these other pieces because there's just no humans to do the work. So the point is like, this may seem cool, but it's table stakes. You've got to do it. I think customers will expect search to just do magical things that never did before. AI will teach us all that any product should be able to turn any amount of unstructured data into the answer that we want into the structured. Is it perfect today? Are these prompts perfect today? No, but customers are expecting it. They're expecting AI to make onboarding much simpler. Right. And they're expecting your product to do the work of more and more add ons and point solutions. So the last point I'll take away, you can, you can, you can fight this and you can say, you know, my competitor says AI does this and it's stupid, or my competitor and but you know what, our customers are smarter than ever. But they're being promised this and they're going to make that bet more and more. They're going to make the bet on the vendor that can deliver this level of automation, performance and AI, so you better, you better keep up. Everyone kept and the customers are being told by your competitors. So easy. We can have this up and running in 15 minutes. Okay? You can lay off a thousand folks in your contact center in a week and this will be better than them in eight days. The promises I hear out there in the world, they're not possible to achieve. But I think if you fight this trend, you're just gonna lose deals. You're just gonna lose deals. And the customers are expecting this is easy. It's not easy. There's gonna be bumps. It is even possible that Klarna may have to bring back Salesforce and Workday. We'll see. But they're still going for it and their customers are being promised it.

So you gotta do it.

The good news is you have to build all of this. You have to build all the automation. You have to be able to turn everything unstructured into structured. You have to build a, you have to build a suite now you have to do all this early. But the good news is you get to do it with less money. That's the good news. Good news because you have to be much more efficient. Now, maybe we don't have to be any more efficient than we had to be a decade ago, but things got a little easier going into 2017, 2018, and then things were nuts, you know, in 2020, 2021 too. Not just for startups and unicorns, but for public companies. So look at Monday's latest chart. In 2020, they 2020, they were able to have 9 is 25% margins, which still aren't that bad for a startup. For public company, it's not great. Look at what happened in those three years from 3 to 28%. This is their, this is their operating margins. Not quite profitability, but close enough. Okay, so this is the new normal. The new normal is that as you get to a billion an error, you're going to have 20% operating margins. And to roll it back in time, I think the new normal is 700 people at 200 million in ARR. So you can kind of roll. If I, if I can only have 700 people at 200 million ARR, how many can I have? 100 and how many can I have at 50 and 40? And these are just levels of efficiency that if you've been doing it for a while, this may seem like, oh, I know what to do, it's old school. But for folks that have been doing this for eight to 10 years or less, this is just a new World you have to do the question is not growth or efficiency. The answer is yes. And the simple answer, whether you look at any analysis the public markets, whether you look at any VC report, Bessemer, Maritech, who see or others, the simple answer is growth is twice as important as efficiency. But they're both important. Rubrik is one for the ages, one of the best evers. But 5XY, it's not profitable. It's not profitable, it's not. So you're getting hit. You got to do both. So that's the good news. You got to do AI, you got to go multi product, you got to do automation and you got to be much more efficient than before. This is, this is. And if you didn't sign up for it, we offer partial refunds out front. It's okay, we understand, you can just go back out, say I didn't know I was signing up for such a hard journey as a founder, tell them how many hours you've been here and they will give you a prorated refund out front because we understand it's harder than you thought. Okay, related point and I made some of this but I really want to get into it for SAS and B2B. What does AI mean? It could mean a lot of things. You're going to hear fun things you heard from, you know, you've, you heard from GitHub and Ramp and others. But I think for most of us what it really means is automation on steroids. Okay. Automation is nothing new. People have been trying to automate workflows in a lot of ways. There's great apps like Zapier which have been doing super cool things. You know, we, we use it. You probably use Zapier or Workado, that's here Others that do all this stuff. This is sort of now they all have AI. But you know, two years ago it was amazing. Automation pre AI. What customers just want is that and 10 times better. And they're being promised it and they're seeing it from Klarna and they're expecting it. And this is how I think about it and this is why. Why is massive spend being unlocked? I mean maybe, maybe it's obvious, maybe this isn't helpful, but why is so much money going to Google Cloud and Oracle and IBM and Azure and all of these others? It's because there some of its experiments but they're expecting massive amounts of automation. That's how you get efficiency in the enterprise especially it's efficiency in the SMB. It may be that there's no humans to do the job, but it's about efficiency. And so we're expecting epic things from automation and that's what you have to deliver. Your job, I think for AI at a strategic level for you is look what was possible two years ago and how can we make this two or three or four times better using AI? It's that simple. What automation exists and I'm sure Zapier and Work Auto, that's all they think about constantly is how can we make this better? Because we're already doing it. This is what customers expect. And I do think this AI bubble is going to pop. Okay. There's just, it can't consume billions and billions of venture capital. The cloud leaders cannot grow at rates we haven't seen in years.

Forever.

Like you just can't grow. Azure just can't grow 28% forever. At some point it exceeds the entire, you know, global GDP of the world of the planet Earth. You've got to, you know these numbers. At some point you have the large of extreme, extremely large numbers. But even if the spend goes down, customer expectations will never go back. The genie will never go in the bottle. They are expecting automation to be 10 times better than it was before. And I just picked out a fun example from, from Gorgeous. But I think Intercom and Zendesk and probably Salesforce have the same thing. But just look at the flow. When you onboard a customer, first you buy their core product and then there's Automate and you just dial in how much of your support you want to automate. 20%, 30%, 100%. You know, five years ago it was zero. You know, two years ago it was some experiments. Now there's folks sliding and there's sliding all the way up and you can see a slider. They have a handful of customers doing 100% and the average customer does like 20 something percent. But this is what they're, whatever your product does, they're expecting in essence a slider. For how much of, how much of whatever I do cannot be fully automated. No humans, no work. That's the promise. And you got to deliver it. So I think this is the meta theme of automation everywhere. And maybe the last point and then I'll wrap up and get, and get Andrew and stay for this one. This one I think multi products now the norm and you gotta start earlier. And we had Parker Conrad from Rippling come to SAS Europa watch his Saster video about the compound startup and it was great. And he talked about how when he did Rippling in a way it was Zenefits 2.0, he started with five products or seven products instead of one. Now he did raise like a billion dollars. So most of you can't do exactly what he did. And he's very honest about that. He's like, you can't do all of it. But the more I thought about, the more you got to do as much of it as you can. You've got us both be focused and be multi product earlier because customers are expecting this. They are expecting every product to do more and more and more. It's the single product ones that are just getting to 5 million, 10 million, 20 million and then just hitting a brick wall when they're not multi product. It's just what you got to do. And I just pulled up too. You know, I don't think the point solution is dead. We all start with two founders or four folks or five folks. We can't all build a suite on day one. So I don't think that Point Solutions literally did. But look at Datadog in Canva, look at the suites they're building. And I don't know the full answer, but I would challenge all of you to just think, where can I make another big bet? It's gotta be big. Where can I make a second big bet? Even as you approach 10 million, I think it's the time and maybe you even can think about it at 5 million, 4 million. Just think, where can my next big bet be? Because customers are expecting this hyper functionality. Last point and then I'll break and get Andrew who will do the work. So I just want to leave you guys with one metric. We did a little experiment this year. We for the first time ever, all the poor folks had to do booth duty. And some of it was all outdoors. Last three years we invited 813 AES and SDRs that did booth duty to see if they wanted to come this year for free. On us, completely for free. Hobnob with the best of the best. Meet the best VCs and CEOs. I spoiled it. But how many you think took us up on it? 4.

4.

4. It's tough out there guys. This room is, I guarantee you is more ambitious than before March 2020. More ambitious. I love it. But I worry when only 4 of 813 wanted to come at no cost and turn their booth duty work into meeting the best CROs. It's so easy to get a job here if you're an AE or an sdr. Just come up to any CRO here, give them the world's best pitch. You're probably gonna get hired today. Probably any CRO here is hiring, but only four wanted to come for free. And we tried to target folks only in the Bay Area. So something to think about. So my last point is, and this could be Klaviyo2. Go find your Monday. Monday. A billion in arrow. 34% growth. 110% NRR, 22% free cash flow. We're going to see very similar metrics from Klaviyo. Go. You can't be Monday, but go find your piece of Monday or Klaviyo. That's the path forward. So thanks, everybody. Let's give me just a break. I'll bring out Andrew and we'll go next. Thank you.

What's the core curriculum of Klaviyo? You have to know our customers. We're here to empower the world's creators to own customers their destiny. How do we allow each of us to be ourselves at Internet scale? A scalable brain and a scalable voice. That's what Klaviyo is. Every consumer business should be running on top of Klaviyo. There's so much more that we can do at our core. We're builders. If you come to Klaviyo, there's a lot of future founders here. Entrepreneurship does not equate to like, you have to run a build business. It's everywhere. Humility matters if you believe in what you're trying to build towards. The rest is kind of just noise. The ambition is still big. Marketing is just the first interface. There's plenty of other interfaces between a business and a consumer. We are 1% done.

All right, let's welcome Andrew Bajacki from Klaviyo. I went the wrong way. Have a seat.

Oh, I'm over here. All right.

Thank you for coming.

Yeah, thanks for having me.

So I want to. There's a lot I want to do. I guess we have a few minutes, so that's good. But let me go back a little bit in time to frame it for the audience. So I think this is the second point. I think Klaviyo is the most awesome SaaS company you don't know enough about. I know it's hard, but. And I'm gonna walk through it. Everything's pretty epic about Cleveland. Nothing's perfect. And I know you're gonna be direct about challenges. Let me step back. Maybe the thing that's most inspirational to me is in 2018, I invested in a company that Andrew knows called gorgeous. They're at 10k and MRR 10k in MRR. They're a good partner. And I got, I knew that this sort of the e commerce B2B ecosystem. I invested in Algoli in 2014 and they went into. So I knew it. But geez Louise. After I invested in Gorgeous, everything I heard was Klaviyo people love. I've never heard more customer love than Klaviyo and I want to dig. The number one thing I want to dig into is why? And I know Andrew will be honest and analytical about it because Klaviyo, I'm going to oversimplify. Andrew, you can, you can explain in your own words, but Klaviyo is email, SMS and more marketing for E commerce. That's maybe to a bread and butter look and tell us more but people love it like grab you on the street by your neck kind of love when like I don't know that email marketing is the newest category. Is it?

It's been around a bit. Yeah.

So I want to dig into it, but I want to. Before we get there, let me ask you a meta question and I know this is going to feel like an analyst call. I apologize.

Right.

Because I. But the slides are good, right? 900 million ish of ARR growing, 35%, 112% NRR from SMBs and 14% free cash flow margin. Is there a downturn?

Not for us. Yeah. So I mean. Well first, I mean for folks that don't know, so what does Klaviyo do? So we, we offer marketing and importantly customer Data software for B2C businesses. You know, when we started, I was always fascinated by this idea of if you think about how businesses interact with their customers, you know, in a B2B world, which is what I'd grown up before, you're used to this model of you automate a lot of stuff, product led, but fundamentally you still, you have, you have sales folks, you have customer service folks. So there's always this like, there's this person that can be the intermediary between the customer and the business. In consumer, that just doesn't work. You know, if you go buy a pair of Nikes, nobody's calling you to find out like how you like them. So everything has to get automated. And when you talk to those businesses, what we realized was, my gosh, like they really want software to deliver like an awesome experience. And the reality is you have all this consumer data, all this first party data, but a lot of people just weren't doing anything with it. But there was no software that did that. You know, when Klaviyo started, the status quo was, okay, well I'll use this spreadsheet. I'll try to segment out my customers. I'll upload that into, you know, my newsletter tool. And maybe if I have time, like, I'll try to send out some like, segmented campaigns. And so the reason that we got all this love in the early days was everybody said, look, I have all this information, I just want to put it to work. I want it to be connected to, you know, the other software I use. And we actually asked them like, well, what software matters? They all said, marketing. And so he said, okay, well that's what we're going to do then. We're going to build the one kind of platform that allows you to store all that data, put it to work in marketing. And then for the first time, people, you know, all of a sudden they're like, well, I have this idea. I've, I don't know, I have all these customers that have bought, you know, jeans from me, but they've never bought any of my shirts. And we're like, yeah, that's fine. That's like a five second thing. And they were like, oh my God, that used to take me weeks. And so all of a sudden that was easy. And then we layered in other stuff that I thought was kind of almost like gamified a little bit. But if you remember back in the day in marketing software, and I still get this question now, folks come up and say, oh, so I know you do something around email. Can you tell me what a good open rate is? I'd be like, open rates? Who cares about open rates? I mean, don't you care about dollars and customers? And so we just built in things because we had access to all the data to say, actually we can just tell you what if your marketing is working or not. And what was crazy to us is for a lot of these B2C businesses, they were used to this with their advertising. You know, if you run ad campaigns on, you know, Google or Meta or even Amazon, you know, they're like, yeah, of course I know what the ROI is. But a crazy thing was when they were doing, you know, first party marketing, they had no idea. So I think a lot of that, you know, initial customer love was just, here's some basic pain points. And we're like, all right, we're just gonna totally solve these and then that's what'll build the, you know, groundswell.

So to, and to oversimplify the 10x feature in the early days was being able to. Almost to. Almost to be able to automate segmentation.

Yeah.

We had this belief that like we knew the people were there doing it in a database or a spreadsheet or all this kind of stuff. And we literally built up, we had Klaviyo.com recipes, which was just a bunch of ideas on segments you could run. And we're like, if you have this idea, like if you think one of these is good, you can sign up, connect your data and we'll do it in like two minutes. And that was just a major lead gen, you know, and all sorts of people just loved that. And you're right, that was like the first killer feature.

I mean, if I'm going back in time and I want to talk about the present in the future, if I'm going back in the early days, probably mailchimp's the biggest competitor back then. Right. And cool app. But segmentation is rudimentary at best. Right. Maybe you have to, you could create like groups or manually do some sort of pick list and maybe do a little geo. But to be able to connect that to your data and it was probably, maybe impossible today, but probably then was literally not even on the roadmap.

Right, Right.

So, yeah, one of my favorite things to do as, you know, as a. Every time it's like studying a product category is just go use whatever people are using and try to do, you know, whatever, like the kind of, the more bleeding edge things and just see how hard they are.

Yeah.

And it was always fascinating. Like, it was like, oh my gosh, it's like, right. It's barely possible. And we said, okay, well if we just make that one thing a lot better, then I bet people will love us for that.

Yeah. And I mean, SMBs just, they don't have anyone to do this.

No, there was, I mean, the other thing too, the other thing we learned too, about consumer businesses, you know, interesting. Like, you know, we talk a lot about like low code, no code as kind of productivity. But for a lot of consumer businesses, what we found is, you know, if you're especially, you know, let's say in, let's say if you're starting, you're going to spin up a new product, a new retail product. A lot of those folks, I mean, they typically come from a product background. Hey, I'm good at either kind of like sourcing products or manufacturing products or maybe I'm good at, you know, marketing. I have some, I'm good at building a brand, but nowhere in there is there a Developer. So there was always this sense that, you know, well, what does it mean to do this without any developers? I actually remember, like, when we, when we first developed built Klaviyo, we had built all these APIs to connect data in. And, you know, like, every good app, we're like, yeah, we built APIs, I mean, now just go get your developer and like, you know, connect into our system. And we actually very quickly realized there was this great, not just like product play, but actually like almost go to marketplace of, oh my gosh, if we go build the connector. So at the time we had this idea of we'll kind of be the, you know, the zapier. Like, we'll kind of bundle some of that functionality in where we'll have the connectors and we'll do that for all these major platforms. It'll save our customers so much time. And so it became a big competitive advantage. And then actually we were able even to pivot that into, hey, we can actually go work with those platforms, you know, and show them what they're doing. I bet they'll be interested. I bet they'll help us drive more demand. So there was this awesome thing that happened in the early days where it gave us better product market fit in terms of, like, helping populate our database with all sorts of stuff as well as give us some obvious partners to go work with.

So, and I kind of had this on the next slide. But you start off as a database build a bunch of APIs, trying to be sort of agnostic to where. And then when did the Shopify hit? Like, when did that just blow everything else out of the water?

Yeah. So like, when you go back in time, when Klaviyo started, we actually, I mean, we launched on hacker news like everybody did back in, you know, early 2010s, and, you know, when we initially were trying to sell to other startups, we're like, oh, yeah, hey, you've got people signing up and you want them to convert, like, we'll help you with that. You know, I had a friend that had a Shopify business. He was selling quilts on the Internet. I was like, okay. So, you know, it was built a pretty sizable business doing this. I said, hey, so I think this might be applicable. Like, you've got a lot of repeat customers. You know, you've got all these different patterns. We can help you. Affinities, all this kind of stuff. And he's like, oh, yeah, totally. I've been looking for something like this. And, you know, I remember, you know, we so we said, okay, well we'll help you. You know, basically again, he was not very technical. And so we said, okay, well, huh, that's interesting. Like what do you use to run your business? He said, oh, I use this thing called Shopify. It's kind of new. I said, okay, that's great. Like, well, how are their APIs? And thankfully they had like, you know, we had spent so much time looking at applications and we found all so many piece of Software. Like now APIs are kind of like a given. But back then it was not like, you know, it was all very, but they were very, very developer forward. And I remember we, you know, we, you know, we helped build a little connector with them and then I realized like, my gosh, like there's actually a lot of these Shopify business, hey, maybe there's a thing here. And you know, after we found the first five, 10, you know, kind of other Shopify brands, we went back to the Shopify team and said, hey, I think we might have something here. And I give them a lot of kudos. They, you know, there, I think a lot of companies, when they do, you know, developer partnerships, a lot of them are kind of like, well, what's in it for us? And you know, maybe, you know, it's sort of, it's a little bit of an afterthought. But that wasn't the case at all for Shopify. Yeah, they were like, no, no, we really want people to build around us. And when they had seen what they built, they said, oh, that's awesome. How can we help promote? And we actually found that with a bunch of other, you know, software platforms. And that was actually a big part of how, you know, you think about how you get the initial flywheel going of like getting folks to, you know, drive awareness. And that was a big part for us is we would go to these platforms, you know, the Shopify's the stripes of the world, and say, hey, we built this thing. We think it's actually great for your customers. Will you help us do some co marketing? And you know, obviously that took off.

Yeah, I remember back in the day, a major platform, we spent like two months being a launch partner on their platform. They said they'd help us market. And then I asked later, what, what happened? They're like, oh, you didn't see the tweet, right? That's what I got was one tweet from a junior product manager. That was how much push we got after what seemed like a huge amount of engineering effort at the time.

Totally yeah, no, we've spent so much time with different. I actually think there's. I think it's a. It's a competitive advantage if you're building a product, if you can get other folks to build around it. Because, I mean, now this is happening to Klaviyo, where we have folks building in and around our ecosystem. And, you know, one thing we talk about a lot as a team is our goal is not to, you know, monetize those. Develop, like, develop. Our goal is, like, we want to build a bigger and bigger ecosystem. And actually what we're finding is, like, folks that build around. It's like, we now have a lot of customers are like, oh, yeah, I found out about you because this other person was talking to me about what they built, and they said, it works really well with you guys, so we might as well come over. So there's. There's these really great, like, flywheels that you can build, you know, if, you know, if you invest in that, you know, for your company.

And I should know this, but today you're not monetizing your partner ecosystem at all. There's no. There's no toll to integrate at all.

Yeah, no, no, no. We're definitely. We are an extension span mode there of, like, I'll give you an example. I was talking to. I was talking to somebody earlier this week who, who's built a, you know, you know, if you want to advertise on streaming Networks, there's now APIs and everything to do that. And so they built a product on top of Klaviyo that's like, hey, you've got all this data. We can help. You guys can provide the targeting and we'll provide the distribution and say, like, wow, that's amazing. That's awesome. So now we just look at as a big competitive advantage where if somebody says, hey, I'm thinking about, you know, my customers, my marketing. Obviously, we do a couple of the channels, but if you want a solution for TV advertising, it's like, well, we don't have anything for that, but now we do. Right? That's the, you know, the power of.

An ecosystem is there, since Shopify was such a boost. And I want to come back to that early story.

Yeah.

Are you using that experience to lean in even more with your partner team? How are you thinking about that at your stage? Like, how are you incenting your ecosystem?

Oh, yeah. So, yeah, I mean, we think about it a couple different, you know, different ways. I mean, there's. There's folks that are, you know, if you think about Klaviyo, we like to think we're the, you know, brain for a consumer business. So we've got all this data. Like if you imagine you could measure, imagine you could remember everything about all of your family, your friends, every last little detail, right. What they ordered for dinner when you went out last week, right. Did they like it or not? I mean, imagine you had all that granularity for everybody in your orbit. Like, that's really powerful. So how do you put that to work? So we think about both. There's folks building applications on top of Klaviyo. So we really try to incentivize that and give people patterns on how to, you know, how to get distribution and, you know, grow into our ecosystem. And also additionally we found is like for other verticals. You know, one thing I've always found about, like when you think about businesses that focusing on B2C. Two thirds of the, you know, global, the US economy is consumers spending money with businesses. So it's a massive part of the economy and we're helping those businesses. You know, as we think about other use cases. So like not just retail, but services, hospitality, et cetera, what we're finding is there's actually that we can do the same thing we did with Shopify. There's lots of opportunities where these other folks are effectively they're, you know, they're sort of there, you know, where you book your reservation or where you buy something. But then they want to, they want to partner up with somebody that can help put that data to work so that those folks like, you know, it could be the guest of a hotel. Like, you know, when you go to a restaurant, like everybody's trying to make those experiences better and so that we can naturally partner up with those folks. And I think actually like the Shopify has become a great template for how we like to go do that with, you know, other software companies.

Yeah.

One question I ask folks when they start to get the Klaviyo scale or bigger is can someone build a hundred million dollar app on your platform?

Yep.

You built a billion dollar app in part on Shopify. There's others, there's multiple folks that have built nine figure apps on Shopify, Salesforce. It's a moment, right? It takes scale. But is it, I don't know, is it possible to build 100 million? Could I be all in on Klaviyo and get to 100 million today or in five years?

Oh, definitely. I mean we.

What's the cheat code? If I want to do that what's the cheat code to getting to 100? What could I build for Andrew and team so I can get 100 million from Klaviyo?

Well, I think it's, you know, like, good, good product development. I think actually one of the things we've tried to be maniacally focused on over the last 12 years since we started, just stay really close to your customers and the kind of the problems that they have. And we go back to this, hey, I want to do more personalization, that being the big burning thing and saying, okay, we're going to figure out how to solve that. If you think about Klaviyo, we're expanding on a bunch of different directions. You know, we're known, you know, most of our customers today, they're mostly small businesses. There's now a growing part that are enterprise businesses, but that's an area that we're growing that we, you know, we're looking for help. And like, there's definitely, like, breadth in terms of the functionality we want to offer there. You know, we're also growing internationally. There's a lot we're doing there. And then we're also growing out, you know, into other use cases in other verticals and use cases both, meaning beyond just marketing and messaging, how else can you improve the customer experience? And then obviously beyond, you know, where we have a lot of customers in retail, like, what does it look like beyond that? So I, you know, for our app ecosystem, my gosh, there's so much white space there. And something that we talk about internally, we can't, we know we can't build all of the functionality our customers want. You know, our, our roadmaps, our backlogs are very, very, very deep. And that's actually why we're so invested in an ecosystem is like, okay, well, if we want to get to all those things quickly, we have to enlist like, others to help us do that. And the nice part is we don't think of this as zero sum. Like, we think there's. There are already. We have businesses that are doing tens, if not close to $100 million that are great partners of Klaviyo. And I think, I mean, I think that's right. The mark of a real platform is can you have multiple businesses like that that can get to 100 million or even a billion?

Yeah, I think it takes a while to figure it out. Right, but, but I think when you have that, then you can get the whole room to build just on Klaviyo.

Right.

Instead of doing 100 applications so let me just. There's a couple things I want to talk about. I think we hit this the old days, finding your ICV a niche. Although two things I want to touch on just before we go past it. That's a funny one.

Yeah.

I'm a big fan of like taking pictures along the way, although I wish some of them would.

You got to take pictures private.

Yeah, yeah, yeah.

Just one thing on this, because you brought it up because we the help from Shopify developers in the early days. Now Shopify is a very different company today. I don't know how big it was back then.

Right.

A 50th of the size probably. Right. But you went to them, they didn't find you in the early days, and they helped boost you in the early days because that's what everyone wants, right. They want to be able to find someone at Klaviyo or Shopify or wherever. And it's hard. Right. So what. Tell us just a little bit about that story, the help you got and why it worked. Like, why did it work?

Yeah, yeah, yeah. So if you're, I mean, you know, when you're, when you're building a product and you think about like, okay, well, how do I get more distribution? How do I go to market faster? You know, for us, I think great partnerships, you need both, Both. Both sides have to. There has to be something in it for, for everybody. Right? So we often think, like, I often use this, like rope. Like if you're gonna keep it tight, you need both sides but pulling on it. So I also think, you know, when you're building partnerships, you should expect to have to do a lot of the legwork yourself when you start. I think sometimes it can sound good. It's like, yeah, we'll kind of, we'll discuss this. We'll agree in principle, and then magic will happen. Yeah, it's like, yeah, that's not really how it works. So to give you a little insight, like how it worked for us, you know, I mentioned, you know, we found. We realized it's like, okay, Shopify merchants were great, great fit for what Klaviyo was offering. Awesome. We spent a bunch of time, you know, with our product building in, making it easy to import data, export data. Right. So there's kind of bi directional sync. We built a bunch of, you know, all kind of basic stuff, like kind of, you know, templates so that folks could get started really quickly. And then we, we.

I had that.

We had this point of view, my co founder and I, that if you could get 10 or 20 people that were just raving fans. Like, if you go to any company and you say, hey, I've got 10 or 20 of your customers, and they absolutely love our products and it's truly genuine. Like, here, here's the list. You go call them up, go talk to them. I mean, everybody's going to stop and look at that. And so we just had this belief that if we go do that, then when, by the time we went to them, rather than saying it's like, hey, you should trust us and believe in us, we say, look, don't take our word for it. The proof's right here. And so that's actually the way it evolved. And actually, you know, like, it gave them a lot of credit for being very open to developers in the early, early days, getting to the first couple dozen, a lot of that was legwork that we had to do. Now, once we got to that critical mass, we then started to solve. You know, I had this. I had this thesis. We had this thesis that my bet was, like, at Shopify, there were folks that were calling into their support team and asking them questions like, hey, I'm curious about how to send an abandoned cart. Or, hey, I have this idea for personalization in marketing. What product would you recommend?

Yeah.

And that this was like a real pain point for them.

And that's what everyone wants, is to be that vendor.

Right, right, totally. And so we just had this belief that they had to answer that question. And so everybody's like, if they didn't have a good answer, I mean, imagine being the person that's, you know, doing customer service, and you're like, I don't know, I'll get back to you. Like, that's not a good experience. I don't want that. So we had this belief that if we could just say, hey, look, we've got these great. We've got these super fans, and then we could share that information back to the Shopify team. Literally, the folks on the front line, their sales team, their customer support team would say, oh, yeah, like, this is great. In fact, what it turned out, by the time we had started working with Shopify partner team, I think they told us they're like, oh, yeah, we already have our customer support team recommending you because they realized it made their job easier.

Yeah.

So once you get that kind of. You've got that momentum, then that the fun things start happening. They said, okay, well, hey, is there some way that we can, you know, can we market you out to, you know, Shopify had was very early to this idea of like, we're going to have an app store. And they said, hey, we've got to promote some, we've got, we've got to, we've got to suggest some apps. So might as well put K. This is one of the ones that to be, seem to be working. And that was great. I think that really helped with like. And then we would do things where we, you know, we do small events together. But I think it all starts from the, you have to get in there and really make sure that, you know, work backwards from what that company wants to see. If they, you know, if you bring to them 10 or 20 customers that are just super fans of your product, of course you're going to take that, you know, that call. And actually the reverse is happening to us now. When I talk to our customers, one of my favorite things to do is, hey, what are the other apps you really like? And so you mentioned things like Gorge. That's actually how I came to know them, is like, oh man, a lot of people are using this and love it. There must be something good going on here.

Yeah.

And that's a great way to, you know, figure out where the, what the partners are that you want.

Yeah, we think, you know, sometimes as founders it can be frustrating or like, oh, Shopify or whomever has favorites. Right. But you got to look at it from the other side. Like they need to make those recommendations. They want to win that deal. And if they need marketing or SMS or something, they can't give you a list of 80 vendors. Like, that's bad. That's a bad. So the simplest thing is to say use Klaviyo, like to have a 500 an army at Shopify. And it's not just that they're favorites, they're solving a problem. Right. And you, there's only going to be maybe two. If you're lucky, they'll mention two names, right?

Yeah. And definitely on that, you know, on that point, like what I found when working with other companies on partnerships, they're actually, well, the good companies actually don't really care about the economics and the finances up front because they're like, look, if this, if there's no fragment fit, if our customers don't love this, what's the point? I don't, I would never want to sell them something that customers don't love. So actually the reality is like, if you want that kind of, you know, favorite, you know, status, the best way to do it is to just bring the proof points and actually What I've found is all the great product companies are actually like, yeah, those are the companies we want to do business with, you know, versus one where it's like, there's some commercial arrangement.

Yeah. Yep. So some. And then I want to fast forward to the present, the future. But some interesting things I Learned. What's the 10x feature? The 10x feature was in part this auto segmenting the base. So it just worked elegantly, right?

Yeah. The two things that we thought were the killer features for us was this idea that, hey, if you have, you know, I remember walking into, you know, our customers, you know, offices when we meet up with them, I'd say if you have an experience for a customer you want to deliver, like, there's. You have an idea of like, I know I have a pocket of customers that look like this and I want to build creative and design that looks like this. I said, no problem. You can do that. And you can literally do that. If you have that idea on your walk to work by lunch, it's done. So, like, that was killer feature number one. And then I.

Lunch, it's done.

Yes. Right. It's. People were just like, boy, that usually takes weeks. And I was like, yeah, with us, you're going to be done, right. You're going to be experimenting. The second thing was I've always felt that, you know, in consumer land, we're better at like, gamifying things. I think sometimes in B2B land, we forget that using software should be fun. It's not just supposed to be a grind. And so for us, one of the things was this, this idea of being able to measure how you're doing and show, you know, the marketers that were using our products in dollars. I remember there's. We made this decision about, you know, a year or two in. We said, you know what, we're not going to show any more open or click rates. Like, you have to. You have to go click 5 times to go find them. All we're going to show you is the dollars and the number of the amount of revenue that say that campaign you just sent drove. And almost overnight people were like, whoa, that was that.

That.

That made $50,000? Yeah, well, that's. That's better than the one I sent last week. Why is that? All of a sudden? They were way more interested and was fascinating. I started to talk to folks and they're like, oh, yeah, we set a goal for our team. They need to draw. I think somebody said they need to drive 100,000 in Klaviyo. Bucks. I was like, okay, well we can come up with any clipbox. But like that's what, that's their goal for the next month. I said, okay, we're really onto something when people start to set, you know, KPIs and goals around things. So that is really, that measuring that really made a big difference because it was, it was so normal in advertising, but it just wasn't a thing that people were doing, you know.

Yeah. Directly ties to revenue and E commerce. Directly ties.

Right, right.

And it's, it's hard to lie or fake it because it's in klaviyo.

Right, Totally.

Yeah. So and then one other thing from the old history that's super interesting and that's also the power of going deep in a vertical rather than horizontal because if you're horizontal, you're never going to build that, are you?

Right. It would be much more challenging. I mean, the other nice thing is one of the benefits we got of. I mentioned this. Like we built the data connectors.

Yes.

I often find like if you build a set of APIs and there's a set of, there's data that you want integrated into your platform. I actually, when we started there was, it was kind of poo pooed. It was like you give like an intern that project, it's like something. But we thought it was like, no, this is core to what we do. A nice thing that happened was because we were the ones pulling the data in, we knew exactly the shape of the data. So when it came to measuring things like conversion, we're like, yeah, yeah, that's easy. So sometimes if you're building a very, you know, say you're building a more abstract product, I think it's actually nice to build a layer on top of it where you provide some schema or some context because you can actually make it a lot more valuable to the user. And it's actually, it's not very hard work. It doesn't feel very sexy. That's the kind of thing.

So.

Right. So we did that in that one vertical and we've now taken that same playbook and said, yeah, what is, hey, what does the equivalent look like, say for, you know, restaurants. Right. Versus retail? And it's like, yeah, it's all, it's all very similar and that, that's going to help us, you know, in that market.

Yeah. Maybe even thinking about it, the fact that instead of click through rates and open rates, we had dollars probably also turned the marketer into a hero. That's probably another reason They're a super fan. Not only could I do it in an afternoon, hooray, but now to my. My boss, the owner not to see. Probably just the owner, not even the CEO. Yes, I can actually. I'm a hero because I brought in another 118,000. I didn't just have a 41.6%. We literally make them a hero.

That's right.

We designed our whole dashboard to be very strong, screenshotable, because we literally. I know, it's like, okay, at the end of the week, you're going to go show this to somebody and we want that to be a win for you. Yeah, yeah.

Okay. One small thing about the past and then I know it's a long time ago and we briefly. You won't remember, we briefly touched on this when we did a podcast during lockdown. But you, you nailed Shopify. You got with the team. You got it, you got in, it worked. You were doing something. Very thoughtful, focusing there. But back in the day, Mailchimp, which obviously exited for 12 billion, pretty good. Pretty good outcome, right?

Yeah.

But they did stumble on that ecosystem. Right. There was friction. There was friction between the organizations. Was that a boost? Was that one of your pieces of luck? What were the learnings when. Because they're bigger. Right. And they did, but they weren't the best partner back then. Is that accurate?

Yeah, I think so. I mean, there's, you know, there's kind of some kerfuffle with like in the, you know, in the.

Well, they got banned from Shopify. I would call that a kerfuffle.

Yeah.

If Shopify says you're no longer allowed to use our platform. That's on the DEF con scale. What is it for a four with the War Games one. It's a four, Right. It's not a little one. It's your band from Shopify.

Yeah. And like I. Yeah, it was. So anyways, our learners now is like, look, if, you know, if sometimes things are going to happen that are outside of your control, that you just, you know, sometimes they're going to be to the good, sometimes they're going to be bad. And like, you just have to do your best to take advantage of those when they do. I mean, for instance, I remember in those weeks we got a lot of folks that came to us and said, I need to move. Like now. It happened not next week, but today. Right. So it was, you know, sort of all hands on deck for a little bit and a lot of like, how do we automate that? But in general, like, I think that was, you know, I'm actually, my biggest lesson from that was just when you're working with partners, it comes back to this. Like, there's got to be in something in it for both sides. And I actually think there's like, the opportunity for things to kind of get, you know, maybe like, you know, go off the rails, like, when they don't need to. Is there. So actually, with all of our partners, we often talk about, like, you know, hey, here's where we're headed. Where are you headed? And like, I often think, you know, you want things, I mean, obviously you want them to be complimentary, but it's actually okay if they overlap a little bit. You just want to understand, like, what folks are after in general. Like, my God, there's so much surface area to get to. Like, we're not sort of not that precious about, like, okay, well, you know, hey, if we're going to color over here, let's just make sure what we're doing is, you know, working together.

Yeah, that's good. Okay. This one's gonna be hard for folks to read. There's a lot going on. But I'm gonna be your Wall street for sure. This is Jason Lemkin from Raymond James or UBS or I don't know. I don't know who the names are today. But there's. Even though this is a. This is an eye chart, it's super interesting because you're coming up on a billion in ARR. You're going international, you're going a little bit more enterprise, you're going beyond E commerce. But you didn't do it super early.

Yep.

So why. Okay, let's talk about. And you're. And you actually didn't. You can tell me whether you think you're multi product. I'm not an expert.

Sure.

But I wouldn't call you radically multi product like Datadog or Salesforce or something like that. You don't have like five SKUs that have almost nothing in combination with each other.

Right.

Even HubSpot. Right. So talk about timing on second act and all this stuff, and then we can talk about some of the details. But what, why and when have you done these initiatives?

Yeah. So let's say a couple things. So the first is when we started Klaviyo and I think there's actually two ways you can kind of found a business. Sometimes you can find a problem and it can be kind of niche. And you're like, I'm gonna go solve this really well. I'm solve for this market. This use case, and that's good, like, it gives you a real, it gives you a toehold, a place to expand from. The thing you just always have to check as you're going along is like, okay, well, how am I going to expand? Like, how do I make sure there's more to this? So you can work it that way? I think that works. Klikio's path was actually a little more the reverse, which was, you know, we had very broad ambitions when we started. We said, okay, well, we're the place that you store everything you know about, you know, your customers. And then we want to be any, any place you want to use that data. We're going to be the, we're going to facilitate that. Maybe we don't build the application, it's just built on top of our platform, but we're going to build a bunch of, you know, apps on top of it and all this kind of stuff. And that was, that was, you know, we had this idea of like, okay, that's what we want to do from day one. But obviously that's, that's so large. You're like, okay, well, where are you going to start? You know, we decided we were going to bootstrap our company. So we're like, okay, we don't have a lot of time here. We got to go find something that's got some product market fit. So the analogy we had in the early days of Klaviyo was, you know, as much like Amazon was like, look, we're going to build this logistics network. This, like, you know, retail websites can be great and we're going to start with books. It's like, books was never the end goal. It was just, look, it's a great place to start. And so that's, that's very much the mentality we've had at Klaviyo is what's the much bigger opportunity? And then how do we just sequence things along the way? So for us it was, okay, how do we build this brain for a business? How do we build marketing, get into that? And even within marketing, we're like, well, let's narrow it down even further. And we asked like, well, what channels matter? And so that's why we start with email. And then we added an sms. Now we, we've added mobile. So now as we think forward, you know, this idea of second acts, you know, my advice is always like, you need to start thinking about second acts maybe a little bit sooner than you're comfortable with.

Yep.

Because I think there's, you know, there's some, you could build some pretty big businesses, but sometimes you get to the spot where oh my gosh, I don't know what to do next. And like you kind of want some lead time to start to build those products, think about them, talk to customers about them. But for us, the obvious, there's actually the biggest challenge for us is there's a bunch of different directions to go, you know, so just purely from a product perspective and being multi product, you know, we started with marketing because after we built this, you know, database back in the early days, literally went to our customers like, what are you doing? You know, database is only as good, you're querying it, what are you doing with the results? Everybody said using marketing, but not. Well, most people said that, but not everybody. And so we always kept a list of what are the other use cases. And just to give you some examples, you know, we, we have a ton of customers that are clamoring for us to either release APIs or product that could help improve things like the, on, you know, the, on your website or a mobile app experience. So literally can I customize how the product feels while I'm using it or you know, I'm browsing a website and then hey, can I plumb this into other parts of the customer experience? So for instance, I think customer service and like all of the different ways of how can I personalize that and measure that? So there's all these other applications and it's a little bit similar to maybe Salesforce or HubSpot that I think Klaviyo can expand into. And again, some of those we'll do directly, some of those you do through partners. So we're very excited about that. In the breadth of, you know, don't just think about Klaviyo just purely for messaging and then beyond that, like, well, it's, this may feel kind of straightforward. You know, we've never wanted Klaviyo to be just a US centric company. We had customers all over the world pretty quickly. You know, I was just talking with, you know, one of our partners in Japan. Japan, our products not even in Japanese and, but they're desperate for like, hey, can you just help us get a little further here? So, you know, one of the other acts we're really excited about is how do we get Klaviyo into around the world. Right. I think one of the things you start in the US like you tend to take this a little more US North America centric view, but the reality is like there are businesses all over the world. So we're working on that. And then you know, finally for us, you know, we started with SMBs. My co founder actually came from enterprise software, but we started with SMBs because we wanted to bootstrap. And now, you know, I made the analogy to Nike's like, well Nike's not a customer of ours yet, but they should be. So how do we work into that? And we know that that's like, that's a very different go to market motion. Some of the product requirements change. So anyway, so we're working across all three of those and then we've got some even like, some even like, you know, crazier ideas of, you know, what does Klaviyo for consumers look like? Hey, if we can as a business, we can help you personalize the experience for consumer. What would you want from us to write? Maybe help you get the most out of your relationships with businesses or discover new ones. So there's actually this like a whole bunch of these things. I think we've kind of done the size and we'd say man, all of these things could be billion dollar businesses. So actually the reality is it's a lot of us for us is like just sequencing those and not getting too distracted. Right. Not getting spread too thin.

Well, you can pick a year, but 2017, 2016 maybe I got my. How big did you think the TAM was for what Klaviyo is circuit today?

Yeah, well we always, I mean we took this like if you, if you looked at, you know, when it comes to like TAM sizing, market sizing kind of stuff, we always look at like try to be pretty constricted because then if you get a big number you're like, okay, that's real. And if it's not, then that's okay. You can always add TAM on, right?

Yeah.

By expanding products or expanding markets. But we already thought that there was like oh, what we were building, just what we were building already could be a multi billion dollar business. Why? We just looked at like the number of customers that were out there. We looked at like the size of their business, how much they typically spent with us and we knew there was already a big there there. It's actually for us, you know, now as you think about it, so there's a long way to run with what we're doing. But I think there's also this bigger ambition of like, okay, well we never, we didn't build this just to sell books. So now like how do we get into these other categories? And so that's And I think we've got the time where we're starting to ramp up those investments and like, you know, those are things that we've got a few years to have them pay off.

But and help me do a little bit of math. This is the non obvious piece. So HubSpot doesn't break out their revenue by products but it used to.

Okay.

Until sometimes it gets more flattering to combine it. But I think they're. My guess is marketing's about a billion of the two because CRMs. Friggin. I mean 90% of this room uses HubSpot CRM and when we started Saster annual it was zero. I mean it's crazy how many folks use HubSpot CRM. It's off topic but yeah, yeah but I think the marketing automation is probably a billion. So you're doing a verticalized piece for E commerce. You're highly attached to one really big platform. Not entirely, but 80% your deal size is smaller than HubSpot's. It's like 5. I mean we're talking in a second about bigger deals but your average deal is like 5k, right? HubSpot's 10. Why are you bigger in marketing? In a niche, a really large niche. I mean it's at half the price. I mean I could build a spreadsheet but why is it so big? Like if you didn't really go through it you might think hey great, great Andrew. But that's going to be $100 million business. How could it be as big as HubSpot if you're vertical at half the price?

Oh okay. So. Well I think there's maybe two things going. One is, I mean this where you just have to kind of look at the economic data. The size of retail just as a category is just massive.

Yes.

I mean it's hard, it's really hard to appreciate unless you're in it. And especially it's hard in a digital world because you can kind of. Actually something I'll do is like when I'm driving around is just look at what are all the stores that are there and like what's their purpose? Like why do they exist now? A lot of retailers moved online but like they're just, my gosh, there are stores for everything.

Right.

And then on the Internet, I mean it's a little bit harder to drive down Main street to see that. But I mean you talk as I talk to customers. Oh my gosh, there are ideas everywhere. Like you know, remember the idea of hey, you know, could, could somebody, could you find A thousand super fans and start a business in retail that's like super possible. You can find, you know, 10,000, maybe 100,000 really committed consumers for some category, maybe some hobby. And it's like. And that's a real business. And there are a lot of those.

Yeah.

So one is. I think the category is very large. The second is, you know, one thing that's interesting about marketing is in a B2C context. So I mentioned this, like, there are no sales reps. You know, one of the insights we had is when you're working as a consumer business, marketing is sales. And it's actually hard to. You have to really appreciate that insight.

Marketing is sales.

There is nobody else coming to make the revenue number if the marketing team doesn't show up. And so if you talk to any B2C business, they'll say, yeah, we own it. We're responsible for this. And I think because our role is as a revenue driver for those businesses, it just has a massively different impact. You know, the value that we can show and therefore what we can, you know, what we can price is different. Contrast that with B2B, where I actually think it's like, you know, if you look at like the, you know, the sales or the, like that, like, that's really the cornerstone. And then the marketing side is. Yeah, it kind of sits in support of that. And it's just this huge difference between those two markets.

And one last piece on it that. And you probably haven't disclosed this metric. And I certainly wouldn't push you as a public company CEO, but I think in core E Commerce, Shopify and Friends. Okay, Shopify, I think you have massive market share. I mean, like more than a HubSpot does. Right. You have. If you've disclosed it, share with me the number. But I think in your core, core icp, it's over. It's over half. I mean, it's insane market share. Right. So how did you learn from that? And why did it not. Typically, I see when you get more than like 10, 15, 20 market share, you have to slow down because it's so much energy to push through the water to get to beyond that. But I don't think it slowed you down to get to this penetration, did it? Yeah, well, numbers don't suggest it slowed you down.

Right? Yeah. The way we've always thought about it, I mean, if you just, if you just go, you do bottoms up. If you go, look at like all the retail, E Commerce, like the B2C businesses out there, we can address. I Mean, there are literally millions and you can, you know how many customers we have. We're still just scratching the surface. So what's interesting is, you know, obviously like, I mean a ton of those businesses are using Shopify. But actually, interestingly, there's a lot of these businesses that were founded before Shopify or just maybe are in markets that Shopify is just more. So there's, there's, there's lots of stuff that's where we, where we can grow. But to your point on like Williams, just take like with within, you know, within some of these like kind of core markets. Right. Shopify, say us. I actually one of, one of the things we believe about software is if you're the market leader and you're topping out at 20% share.

Yeah.

Like you're not. That's not really much of a market leadership position. And sometimes there's dynamics why that happens. But I often think it's like if you have truly built the best product and maybe there's some network effects to it, you should be able to top out way higher than that. Right. Could be 50%, could be even much larger than that. And I think that's one of the things that actually we've talked about for years is as we were growing your point? I think a lot of people say like, oh boy, it's gonna be tough to push through this. But actually I think if you, if you really get the product market fit really snug, then you can just kind of get, you can kind of get to this default status. And you know, that's, you know, I think you've seen loads. I mean this happens a lot more in consumer lands. But I think if you can do that in B2B, where you just become the default very hard like that has all sorts of wonderful. Yeah. Side effects.

It's a good insight. Okay. Just two things I want to hit before we run out of time. Thanks for your patience on this one. One just very tactical and then one, I want to talk about little things like IPOs in the public markets just to help everybody.

Right.

Just a small one. I could have picked a million things. But in terms of timing, you're just really going mid market and up market now. Right. This is your slide. I mean it's great. 23, 86, 50K, 64% year over year. And you're seeing the classic thing, which is even though klaviyo growth is epic, these ones are even Epicure. It's 64%. Right. But it wasn't overnight. And they're really big merchants out there. I mean, Shopify, celebrate. I mean, Harley and everyone, every time they have a Kardashian or SpaceX, it's the whole extreme. Right. Why were you deliberate here? Like, what are the learnings of pacing and timing?

Yeah, yeah, yeah, yeah, definitely. So. So Klaviyo started with, you know, more SMBs. Great. And we kind of had this. We knew, you know, a little because of our background, we wanted to get into the enterprise at some point. So I tend to think of like, well, how do you get into the enterprise? You know, one of the things we didn't want to do is we're not just going to hire a big sales team, give them some territory and say, go, start knocking down doors and like, you know, and like, you know, grow, grow customer. Like, you've got to start with a little bit of like a groundswell. And so that was something we want to be very deliberate about. I think when you're. When you think about the unit economics of your business, especially, you know, your cost to acquire customers, we at Klaviyo, we've always felt like the more you can do, obviously with your products to make it, you know, have better fits, you can do things with, with customers to drive awareness. I'm a massive believer sometimes. I've had folks in our marketing team don't like saying this. Our best marketing is like viral marketing. It's word of mouth. It's people just talking to each other.

Yeah.

So it doesn't mean. It doesn't mean that, like, the market programs you run aren't important, but you want that basis where you've got customers and then partners, you know, so we often felt like when we go to market, we actually, by the time anybody from Klaviyo talks to you or you ever see an ad from us or anything, you should have talked to your peers. You should have talked to maybe some partners. Like, maybe in our case, maybe it might be a marketing agency. You've talked to some of the platforms that we work with. And if they're all saying Klaviyo, by the time you talk to us, you're like, oh, yeah, I'm like, already predisposed. We're like, well down the funnel. Right. So when we were moving into the enterprise, we tried to make sure all three of those ingredients were there. Did we have customers that loved us? Did we have, you know, partners, you know, in our case, digital agencies that already knew some of those folks. And we're excited about us because they usually have a really good sense for what Products will work.

Yeah.

You know, obviously, we've done a lot of work with Shopify and others on the platform side of let's work on this together. So I think when you kind of stack those things up, then more recently we said, like, okay, let's pour a little bit of, you know, sales and marketing gasoline on this now. Now all of a sudden, you know, it's going to work. That plus I think going enterprise, you know, we've started to expand the number of products we offer, so more products within marketing and now some products outside of that. I think there's this very big push in the enterprise to I want all my software to work together. And we've been able to play to that, too.

All right.

There's more I want to talk about. We could do a sequel on the podcast sometime if you want. But one thing, I know you're probably bored of talking about this, but just to help folks here, Klaviyo IPO'd September 2023.

Yep.

The last one was Hashicorp in our universe. So it was. How many months is that a gap? 20 months. Is it closed? Obviously, it's not closed, but a more nuanced answer. There's a, you know, social media is annoying, Twitter's annoying, but how closed is it? What did you learn? What's the real bar and what's the takeaway for founders from the world? It's because it ain't easy, right?

Yeah. Well, I definitely think, like, you know, so I go back to when Klaviyo started in, like, the early 2010s and what the bar was then. I think in some sense, we've. I think that, you know, maybe the ARR number's gone up a little bit, which I think is just too bad. I actually think it should be, you know, if you're 100 million, $200 million business, like, that should be plenty. Like, I mean, you can do more growing in the public markets. You don't need a billion in revenue.

Like, that's Shopify and HubSpot IPO to 100.

Right.

And I think. I think that will come back. So, like. But I think this idea that, you know, hey, it's a sustainable business, that it's a business that can grow durably, that it's, you know, it's got positive free cash flow. I think that's real. Like, I think you have to figure that out. There's probably still some acceptance, like, maybe if you haven't totally figured that out. But I think you need to have a path to, to do that. So I think if you meet those criteria, I think you're in good shape. And actually, you know, just a bit of a meta point. You know, I felt like when we started was a great time to build. It was. It was still maybe a little bit of a hangover from, like, you know, the 0708 recession. I tend to think right now is like another awesome time to be building. It's just, there's just fewer people that want to do it. It's a little, it's comparatively a little bit harder, which means that, like, it's actually less competitive. So, you know, there's actually, you can get more time with customers, all this kind of stuff. So I think there's a lot of companies that can and probably will go public in the near term. But for us, you know, we kind of looked at it as like, look, we knew we're going to go public at some point. I often used to joke with our investors. I'm like, look, like by 2030, we'll have gone public. So will I remember which year it is? Probably not, right?

Yeah.

So, you know, when we looked at it, like, there's just a bunch of work that you have to do, and we kind of decided it's like, okay, we're here, we're ready, like, you know, like all things, like, we have a lot of urgency. Let's just go get it done and then let's move on. And then, you know, folks often ask, like, what's it like being a public company? Honestly, like, it's not really, it's not really any different than, you know, being private. Like, I actually, I think we as a team, actually kind of like the rigor of every few months, you have to show how you're doing.

Yeah.

I think there's a lot to be said for setting goals. Right. Being able to show up and meet them, you know, and then, you know, if you're, if you're a founder, you're naturally going to dream long term. And then you just, you know, I think really good companies can balance between those two, thinking long term as well as execute in the short run.

And just for folks that there's some. There. There was, there was a discussion at SAS Europe and London this year. Welcome to come anytime Europe if you want to. A bunch of folks that were at 100 to 200 million, that were in more impacted categories. You said there's no down there. They're saying 20% is the new 50% that, like, it's hard out there. So 20, if we can just grow 20% in our. What do you. If you had to gut check at having done it at 200 million, what do you have to grow at IPO? Forget about efficiency. I know it's important.

Yeah.

Is 30 enough? Is 25? Can you do it in the 20s? Does it have to be? I mean, HubSpot and Shopify IPO at 100, but they're going north of 50 like you did. What do you think the line is? Just having done the roadshow, if you had to ballpark it, what do you think the line is to have a decent ipo?

That's a good question.

I mean, it's definitely like, you can't be colored too much by what we've seen the last, like, two, three years. So I definitely think it is, it has come back. And I think it's all, it is, it's all relative. Like, I think everybody's looking. So I think you can kind of look at the public market comps. I think there's generally an expectation that, like, look, if you're a smaller revenue base, that you should be growing faster. But I also think equally important is the durability of that growth. You know, actually, I think when you, when you talk to a lot, a lot of folks in the public, it's probably true. The primars too. I mean, everybody wants, you know, what's, what's the easiest path. And so I think even if you're, if you're a 25, 30% grower, but you're like, but I have a very durable model. I think that can last for four or five years in the future. Yeah, I think people really pay attention to that compounding. And so that fact that people know that there's, there's like, there's some, there's some heft to it that's probably equally important. So I think that could be enough. But you, you've got to have a point of view on, like, how are you going to sustain that in the long run?

The fragile revenue isn't going to cut it. All right, Andrew, I have more, but this was great. Thanks for going so long. Thanks, everybody, for being with us.

Thanks.

This was great. Thank you.

Ladies and gentlemen, please take your seats. Our show is about to begin. Please welcome to the stage amit joar, CEO at attentive.

Hello, everybody.

I'm super excited to be here with all of you today. For me, being able to present to you some practical steps that I've seen that enable companies to achieve 20% growth over and over again and compound that has been essential to my career success. My career started 11 years ago in mobile commerce. I joined a company called Braintree, really focused on building mobile APIs to enable payment solutions that you saw for the one click checkout that deliver experiences like the Uber, the Airbnbs, the Wish.com, the Grubhubs. As I joined as COO and CFO there we acquired Venmo to fix the peer to peer problem and then sold the whole business two years later to PayPal. Within those two companies we saw that compounding growth was happening based on building technology that the customers needed and focusing very specifically on mobile commerce. I transitioned after running both those businesses to Attentive where I focused on a different part of the mobile shopping experience. How do you get the right content to consumers so that they can browse the items that are most relevant to them and able to check out as fast as possible? Attentive is focused on mobile commerce specifically on the marketing side. We send SMS and email to some of the best brands on the planet. We have more than 8,000 customers who using the product today. Brands that you'd recognize from retail, e Comm, direct to consumer travel and hospitality, many of the sports teams. And we created a solution that allowed them to focus on reaching their consumers where they were on their mobile devices as they were browsing and looking for more information about their brand. And it's resulted in some tremendous success with for us, right? We've gone from 0 to 500 million in revenue in less than seven years. When you see the standard SaaS business, only 4% of them ever get to even a million dollars in ARR. We also have been in a pretty rare era where 0.4% of companies that are pre IPO ever achieve $500 million in revenue. And we've done this with huge brands that use our platform to send more than 32 billion text messages last year and we delivered more than $20 billion of revenue from those messages where consumers click on the links we have and end up buying by reaching them where they are. So I'm going to spend my time today talking to you guys about how we did it, what were some of the lessons we learned and what has made us success successful. And hopefully you'll be able to take these to your own business and compound on them. The first problem that Attentive really was able to learn and solve for us was build a moat by solving a hard problem, right? For us. Hard problems exist for a reason. It isn't that you only found them, it isn't that they were invented, they are holding back real customer value. And when you look at the problems in detail, find ones that are worth solving and you create a technological advantage by having a solution to it, you can create a legal advantage by having a patent to protect it. And you cement yourself as a thought leader in the space, unlocking that real customer value. So if we rewind to the founding of Attentive, there was something that was very clear. Consumers were more distracted than ever in a mobile world, right. If you think about the behavior on phones, you see people spending less time interacting through traditional media. Tv, radio, Billboard magazine was losing effectiveness. It was harder to track. So how do you reach these consumers where they are? The second thing that had happened was that consumers expectations had changed. They were expecting a more interactive experience. Social media and influencers had raised the bar on what was considered good marketing and good advertising. So there were new influencer sources that mattered and there was an urgency to solve this problem. The urgency came from the brands themselves. Marketers were spending their budget but getting less performance, lower roi, bringing in less new customers and ultimately generating less revenue. So core problem, real issues, no simple way to solve it. What Attentive identified as similar to my career was that mobile commerce was taking over the world, right? In 2014, roughly 15% of mobile shopping sessions were happening on a phone. If you just fast forward four years later to 2018, you saw 40% of online shopping sessions happening on a phone. That's almost a 300% increase. And what you saw with that was this opportunity to reach these users where they were, the payment side had been solved. How do you now deliver messages to them? And today we see at 2024, almost 75% of all shopping sessions are done on mobile devices. Where the attention span of consumer continues to shrink, where it's more important than ever to give them the content they need. So we identified SMS as the channel to reach consumers on these mobile devices. SMS is basically a text message. The problem with text message historically was when you use this as a channel, there's a lot of regulation around it. So brands were scared, there was a lot of compliance. And the standard way of signing up for SMS messaging was that you would give a brand your phone number and then they would send you a message back and you had to reply why? At some point in the future that broke the shopping momentum for any online brand that was getting someone to check out and the person wanted a coupon code moving them off the app, waiting for a reply. Why? Message breaking that flow was Basically the worst thing you could do to conversion. So Attentive spent months and months researching with brands, with consumers, with regulators, how can we solve this problem? How can we unlock the value of these shopping sessions and give brands away to communicate with consumers over time? And what we found was a compliant solution that allowed for easy consumer opt in and opt out. And that was our patented two tap technology. So if you'll see on the screen in the animation, you go to a site, you want the 10% off coupon or the free shipping, this first click, you click on that, it automatically transitions you into your messages app. And then from the messages app itself, you press a pre composed message to send out a message from your device to us. So then we fingerprint the device, we get a message directly from the phone number and now we've met all the compliance obligations, we've given a very simple experience to consumers and brands now have a channel to reach consumers where they are. We win on all fronts. And compared to the legacy solution of reply, why it's fundamentally different. It's in flow, it's fast and it creates loyalty. It's also not an attack vector for the reply Y. When you give your phone number out, if there's someone you don't like, you can hire a service to have that number signed up to thousands of sites that will just keep texting them reply Y over and over again. Something that the carriers really were worried about because it had been happening over and over. And as you think about SMS marketing at its core reaching the consumer on their mobile device, you have to have people signed up for your program to even have a business. So inventing this technology is really the core of what started Attentive. It positioned us as a thought leader, it created a technical moat with the patent and it changed the paradigm of how retailers thought they could engage in the channel. As we move to the second lesson that was very successful for Attentive, we turned our customers into our best marketing team. Very specifically, making those customers successful and telling others about it worked better than any message we could produce from the company. We did this by asking each customer very specifically what their KPIs were, how they defined success. And we made those our own KPIs. We built those into all our OKRs. And we did this across teams, across sales, across product, across implementation, across our CSV team. So ultimately we are all aligned across the entire business that when our customers are successful, we are successful and if our customers are not successful, we cannot be successful. We did this very formally so when we would inform a customer that they were up for a free trial, we would go and create KPIs together. We asked them one very simple question. What level of performance would you need to confidently sign up for attentive at the end of a trial? And we would have them email us those numbers. We were experts in this space. We knew what boundaries to put around them, guide them to an answer that would make us successful. But getting that email, having them write down what success looked like, gave us a benchmark and then send set the boundary point over the free trial of what we needed to hit. So whether they wanted signups, so they had people to message, whether they wanted to generate a lot of engagement for every message they send, or revenue or roi, we knew it, they knew it. There was no disagreement at the end of the trial. And then we built a bunch of reporting that highlighted those specific metrics and we could highlight where they thought the bar needed to be so we could show when we were above it. And you're pretty damn sure we worked hard to be above it as many times as possible. So once we had done that, it was time to get them into the free trial. We've defined criteria for success, they've sent us what would look as success internally for them on each KPI. And we took the first 30 days to give them a crash course on seeing SMS as a new marketing channel for themselves. And we worked very hard on this. We would have meetings every week between our sales team, our implementation team, our CS team and our product team on how do we remove as much friction from the process as possible, where are points where we can make it easier to sign up, when should we show the results of how the trial is going, what is most compelling and what are the ratios we have of people in trial as a live customer converting into a paid customer? Because we've done all the work when you have someone live of finding them, acquiring them, getting them to trial the product we need to get them live and paying. And this became an entire ethos for us as we had our okrs. As we were looking at the metrics for our customers, the other metrics we layered on top were these conversion rates. And what we did at the end then was a very formal free trial review. We had their email to us as a single slide. This is what you said would let you confidently sign up for attentive. At the end, we would show our performance relative to those metrics. They could see the revenue generated, they could see the number of people they signed up, they could see the value of the incrementality of SMS as a channel as opposed to traditional media. And then we had the basis of a conversation of really, what was price, how do we make this work? And that was foundational in the reporting. It was foundational in, in the way we talk to the customers. Foundational into everything we did to start building a base of customers that could set criteria of what success looked like. Generally, those criteria were metrics that made them look good internally. We went out and then beat those metrics, had them convert at very high rates into customers for us, and then shift over and start telling their network about it. The second thing we learned on this path was that as you're reinventing a category or building it out, you need to spend a lot of time in educating, particularly your early adopters. And boy, did we spend a lot of time in education. We actually created what we called Attentive University and we would teach people how to use SMS and text messaging as an effective channel, how to mix it with email, which customers to message, which ones not to message. We actually ended up creating a credential and a certification on the attentive platform. And people would put this on their LinkedIn profiles that they achieved attentive certification. And it became a criteria in job searches over time that people had this credential, they knew how to use the platform. And when you knew how to use the platform, it became reinforcing where people then wanted to have that credential so they could apply to the job. We also highlighted texts we love, which were just great examples. We thought of innovation happening on the channel. Marketers loved to be selected for this. They would get lots of praise from their own population and own peers for having innovative campaigns. We then spent time in our blog of explaining what was happening in the ecosystem and how SMS had more frequent interaction than email, which made it a much more durable channel. The cookies were refreshed even as Apple was doing their privacy moves. And finally, we had formal events with beyond the Message where we had an equal number of prospects and customers coming to our events to hear what our new product release, hear what the results were. Even people who were on competitors were coming to us to see what was coming next and what they were missing out on. These were great opportunities for us to keep our leadership position, but also help our customers become our marketing engine with the prospects around them in that thought leader position. The most important time of the year for most retail and E comm businesses is Black Friday Cyber Monday. So we create a full guide which we call the archive for this specific holiday. Everything you should think about, everything you should do. And this has become the standard guidebook. We have more people download this that are not our customer than that are our customer at any point in time. But those are all leads for us. Those are all people we invite back to our events where we can mix customers and prospects so they see what they're missing out on, they see the opportunity there. We also have events where we do this in person, right? Our thread event is a mix of customers and prospects talking to each other in major cities where marketing is happening and where it is going. We also have a luminaries event where we take both customers and prospects who we think are at the cutting edge, where's the market going? And we do a two day off site with them. Tell us what's coming next, how can we solve your problems? What aren't we doing well enough? Let us be on the cutting edge. And all of this together has created this self fulfilling prophecy, right? We invented the channel, we built the credential, we had a way to show customers that they could be successful with us. They would get promoted eventually for the results that were outside expectations because they were performing better than expected. And they would tell their peers and their friends, attentive helped me do it. They would then move to another company. As they moved to another company, they brought Attentive with them, right? And as they were at bigger companies, they, they told us the problems they were facing there, which gave us a leading edge on what to go solve next. So customers became prospects, prospects then became customers, and it kept fulfilling itself. So in the end, our best marketing was actually our customers telling the story authentically to the marketplace as opposed to us trying to scream it from the rooftops. And the results speak for themselves. We generated more than $20 billion on behalf of our brands by consumers clicking on the links in the marketing messages and buying. And for customers that used us year over year, we've seen their revenue grow more than 25% from 2023 to 2024. The third lesson, and maybe the most relevant to most of the people here, is when you see a technological change, you need to be early, be opinionated and be bold. For us, we want a leadership position in the category that we reinvented with sms, right? So we need to have a perspective and we need to tell people about it and share that as broadly as we could. While we took what we thought was opportunity and made it reality, specifically for us, the opportunity was that generative AI allowed Marketing as a paradigm to change fundamentally. For the first time in mankind, you could write marketing messages at an individual basis. Each person could get a message that was tailored to them. It wasn't just one message for hundreds of thousands or millions of people. Tv, radio, Billboard magazine, email, SMS had always been one message for a large segment of the population that was fundamentally changing. And we knew that AI was going to revolutionize the shopping experience. We told people about it, we published our thesis, and then we got to work, right? We started pounding RVCs for information on AI. Who are the experts? Who can we talk to?

What are the capabilities?

Capabilities today? What will they look like tomorrow? What do they look like a year from now? We went to our customers early and our prospects and asked them, what problem do you want us to solve with AI? And we were actually surprised here. Almost none of them said they wanted us to focus on the cost side. They wanted the revenue side, which was very good for us to focus our attention, right? It wasn't to replace people on their team. It was really to generate more personalized experiences that got more clicks and ended up with more purchases. And as we gathered more feedback from our trusted customers and prototype products with them, we aggressively shifted a roadmap at a company that's doing roughly 500 million in revenue a year. That's a big deal, right? We have a formal financial plan. We had product and roadmap that we needed to ship to hit the goals. But we saw this opportunity, we saw the moment in time and made an aggressive change by moving 30% of our best people onto AI immediately. And then finally we moved into trial very quickly to create this iterative process of learning by trial testing with our customers, achieving the results they want, and again moving back to this place where they define success. We created joint KPIs and then we worked to go exceed that level. And we did that with AI. And what we saw was we were able to seize the momentum, right? One of my favorite quotes is, it's better to be an average surfer on a great wave than be a great surfer on a flat ocean. We're riding the wave of AI creating the momentum on top of a technological change that existed. We were opinionated, it was very distinctive on where we sat in the marketplace. We were also focused on getting customers into the product, testing it, setting these goals, and very, very clearly setting the value proposition of we are going to create revenue for you. We are not cost saving, matching the market need. And as we've put these together, I'm very proud to say We've shipped two AI products right now that are in.

Market for the last quarter.

We have hundreds of customers on these products today paying for it. And when customers are using personalized shopping experiences, we see that they can generate more than 124% incremental revenue. Right? So as opposed to just saying Labor Day sale, 20% off to be able to speak to the user about base or soccer or tennis or whatever their sport of choice is, gets more clicks and it gets more buys. So just in summary, there were three things that helped Attentive continue to achieve its 25% plus annual growth rate. The first was solve the hard problem that really started the business, right? We got people signed up on sms. We made SMS a viable channel. We worked with all the stakeholders, the consumers, the retailers, the regulators to find the right solution. And that created the moat. We also got it protected by patents. The second one was make your customers your marketing team. They're better than anything you can say yourself. When their goals and their KPIs become your goals and your KPIs and you define the success together, they are going to take you with them wherever they go. They are going to tell their peers you brought them success. Enable that flywheel that is more powerful than any dollar you can spend in outbound marketing. And finally, if you see a major technological wave, do the work to have an opinion. And when you gain conviction, be aggressive. You have to be bold, right? Technology changes don't happen all the time. But you position yourself as a thought leader. You position yourself in a place where you can capture value and you position yourself in a place where you're learning faster than anybody else. So with that, I really appreciate the time. Thank you so much. And hopefully you learned something. IT. Sam. Sa. Ra.

Ram.

Ram. Ra.

It.

Sa. Sa.

It.

It.

Ra. It.

Sa.

Sam.

Ra. Ram. It.

Sa. Sa. It. Sa. Sa. Sam.

Ra.

Sa. Ram. It. It.

Ladies and gentlemen, please take your seats. Our show will begin in a few minutes.

Sam.

Sa.

As a courtesy to our speakers and fellow attendees, we ask that you please silence all devices.

Sam. Sa.

Ladies and gentlemen, please take your seats. Our show is about to begin.

Sam.

Please welcome to the stage Raja and Shet CMO together. AI and General partner at Hypergrowth Partners and Emily Tsao, Principal at Salesforce Ventures.

Welcome everyone. I'm Rajan, General partner at Hypergrowth and also run AI Go to Market accelerator. At Hypergrowth, we help AI companies and B2B SaaS companies with marketing, growth and sales. As operating partners, we have over 40 companies in our portfolio ranging from Together AI, Vercel, RAMP, Ritu, and many others. I'm also currently serving as an interim CMO for Together AI, an AI infrastructure company. I was previously VP of Marketing and Growth at Cohere, a foundational model company, and before that at Heroku, a developer platforms company.

Hi everyone, my name is Emily. I'm an investor with Salesforce Ventures. Hopefully you all have heard of Salesforce before. If you're using HubSpot, I recommend you seriously reconsider that decision. And with Ventures program, we've been around for 15 years. We've deployed over $6 billion into over 400 investments. I personally focus on AI and have invested in some companies that hopefully you all have heard of before, including Anthropic, Cohere, Runway Hugging Face, and most recently Together AI, which Rajen is working really closely with. And thank you so much for joining us. I know it's after lunch and everyone's probably getting a little sleepy, which is totally fine. It's okay if you fall asleep in your seat. Hopefully Rajin and I can keep you engaged and entertained for the next 20, 30 minutes that you won't fall asleep. And we're here to tell you a little bit about what we think in terms of scaling an AI startup from 0 to 100. And I'll start by telling you a canonical phrase that a lot of you have already heard of before, which is triple, triple, double, double, double and refers to a well known profile of what B2B software growth should look like for the new generation of B2B software startups. Well, that might be true for the traditional B2B software startup. We are here to argue that this new way, this new crop of gen AI startups actually look quite different today. Over the last 12 to 18 months we've actually seen an increasing number of startups growing at a faster clip, breaking defying this general rule of thumb and to give a couple well known examples that are publicly reported. So I'm not breaking confidentiality anywhere. Hey Gen, which is an AI video generation company that you all might have tried their first products before have been reported to have grown 1-20x over the last within the span of a year. Similarly, a B2B software company in the legal tech space, Harvey, has also similarly grown as 1 to 10 over the last within the last 12 months. I think what these data points are telling us, well, because they might be a little skewed given there are publicly available data, what they're telling us is that the bar for what best in class growth looks like has Also shifted higher. And to illustrate this in chart form, we have here three different companies and their theoretical projections for these three companies over the next seven years or the first seven years of their journey as startups. This is what I like to call the scaling loss of B2B software. You all have probably heard of scaling loss of foundational models for the traditional B2B software that's growing at triple, triple, double, double, which mind you, was the best in class profile previously, it will reach this arbitrary milestone of 100 million ARR in about seven years. And the very simple math behind this would imply that the two generative AI startups that are growing at a faster clip in the initial years will get to this milestone a lot faster. And this Milestone can be 15 million ARR, 10 million ARR, or your next fundraise, or the next time you hire a go to market team. What the higher growth profile or the standard for growth implies is that the timeline to get to that next milestone has also been compressed.

Similarly, on the B2B SaaS cycle, we have seen the traditional B2B SaaS cycle is very predefined and has a playbook very well defined.

Fine.

But what we have seen is that there's a serious timeline compression happening in B2B SaaS. We saw that marketing and growth comes after the initial product market fit and there's already a hungry sales team waiting for leads. But as you think about the new generation of AI companies, these cycles get compressed. And what we see is that there are multiple of these cycles you go through within a span of one or two years. Because the entire landscape, including the technology, the market as well as the product, is in a constant flux. For example, echo here we saw we went through these three cycles within a span of just one year. We started focusing on very focused LLM use cases. We found product market fit and we started growing. And as soon as that happened, ChatGPT came out. And what we saw was that every enterprise wanted to transform their entire business focused on hundreds of use cases versus singular use cases. So what we did was we immediately layered our enterprise strategy on top of our PLG motion. And within a few quarters what we saw was that there are new foundational models coming up. The existing ones are raising billions of dollars almost every month. So how do you compete with that? We wanted to flood the market, so we layered on another partner strategy and used that as a distribution mode for us. And that's how we were able to scale pretty quickly, layering one motion off the top of the other. Within A span of one year. What this signifies is that there's a serious timeline compression. Your traditional B2B SaaS is very different than the gen AI companies in terms of how fast do you move in the market and most of the times your marketing and growth needs to come before even you find real products market fit.

So the agenda for today is I will talk about the reasons behind why we're seeing this shift in this new growth profile and Rajan will go into the tactics of how you can achieve the same as operators and builders in this space. As amazing as the Harveys and Haitians of the world are not seeing that their products are not good or their go to market is not good, they've done some something, they've definitely done something right to achieve this level of growth. But what I'm also arguing is that I'm betting that a part of that growth is what I call experimental revenue and it's not a new idea. I see some people nodding, the audience, investors, some of the investors I've spoken to and some of the investors you will meet in the conference, at the conference will probably agree with me in terms of this concept of experimental revenue. And there are a few reasons causing this, but we're really comparing it against what ARR is. ARR. We like ARR. Air is appealing to us because it's recurring, it's predictable. But in the world of AI, everyone, all the customers are trying or constantly experimenting with different vendors and different solutions and use cases that the revenue you're getting from these customers are not by nature fundamentally recurring and predictable. And there are a few different causes dynamics that are driving this emergence of experimental revenue. And they can all be attributed to this, to one word, which is nascency. Nascency in the applications and solutions are getting built, nascency in the technologies are underpinning these applications and solutions and nascency in customer behavior, which is kind of the three causes that I'm trying to illustrate in this very text heavy table. So first of all, foundational models are enabling a platform shift, which means there is an explosion of different ideas, different startups that are built on easier ways to build products today and software today, which means that customers have a growing number of solutions to be choosing from and underpinning that the underlying technology, the AI models and infrastructure that support these models are also at the same time moving very quickly. Even though anthropic and OpenAI and cohere have been around for many years, the last two years we're seeing progress being made every week Sometimes every day. And what that means, that is the applications and solutions built on top of it are also moving with it. And because of nasiency in these two layers of the AI stack, customers customers are also unwilling to make long term bets, even though they're very eager to adopt AI. How many times are AI is AI mentioned in a public company's earnings calls? If you're not adopting AI, you're falling behind. And what that what ends up happening is customers are trying multiple vendors at the same time where they're paying smaller ACVs that do not reflect their true ability to pay. And all of this are culminating in emergence of experimental revenue. And what that means for investors like myself is doing diligence in companies to figure out where is experimental revenue becomes a little bit more nuanced. But hopefully we're getting better at it. And for founders and builders it means that do not mistake experimental revenue for real growth. And we should think more deeply about how do we find signs of it and ways to combat at that. So investors are kind of similar to sales reps, which a lot of people don't realize. We have to build our own pipeline. Sometimes we bother the hell out of our founders by emailing them ten times a day. And we also have to qualify that sales pipeline by looking for signals. And recently we're looking at signs of experimental revenue as part of that qualification. I'm here looking at four broader buckets which we won't have time to go through every single one today. So I'll just pick on a couple that I've seen more recently to show kind of an example of experimental revenue in terms of retention, which is the most obvious bucket a situation, a scenario of a startup in this space can have really high growth rate, really fast flying growth rate, but the same time also really high churn. What's happening with that is you might have a product that's super attractive for initial users and have that high growth initially, but not enough investment in the product itself to retain the product, to retain the user on your platform. While that's a very simple metric to look at, digging a layer deeper, more recently I've started looking at contract length, specifically the relationship between contract length and churn. Take this AI infrastructure startup for example. Both it has two different products. Product A is a more complex product that doesn't have a lot of competitors in this space. Product B has a lot more competitors. As you can see with the churn rate and contract length. What's happening underneath the beneath the hood is with product A, the team there has been successfully increasing contract length while decreasing Churn, but hit a ceiling when it got to the six month mark. And when customers are presented with a six month renewal contract, they ended up going out to other players, different vendors for an RFP to see who can offer them shorter term contracts. Whereas for Product B, while it's having a tougher time in terms of increasing its contract length and managing Churn, Churn is actually going the right direction because the sales team that's in charge of this product is doing a great job of qualifying its leads, its icps by just focusing on customers that have real use cases that can be put into production. Which means that this product is being used in a live production use case that could be external facing which is a lot harder to rip out or internal facing and driving a lot of roi. And if I can go back a couple slides to the summary slide, the second bucket is usage and use case. This can be the frequency of usage and the quality of usage. And one example that often happens with a prosumer kind of profile and some B2B. Traditional B2B profile of products is usage metrics can fluctuate wildly with new product announcements and feature releases, which is what's happening with this AI video generation startup. Every time we're looking at the weekly active user base for this AI video generation startup, every time a new feature or a new model is released we see an increase an uptick on monthly basis in terms of weekly active users. But that falls down tapers down pretty quickly in the months following. And while the average weekly active user is going up to the right in the right direction, what this means that this company is not retaining all the traffic. All the users that end up on the platform initially when they see a new feature or a new model released because they want to experiment with that new model or feature, what they can potentially invest in is activation methods that happens post announcement in product in platform that retain the traffic that it's so amazing to have after the announcements happens. And if we go back to the summary slide again, you'll see that there are a few other different buckets examples I've included here under usage. And one thing to point out that I'm seeing a lot recently, recently is the last one here, customers using multiple vendors at the same time for different use cases. This is especially true in a few different categories and markets out there. One of them is security for AI. Think about cyber security for the entire AI pipeline. One of our investments at Salesforce Ventures is this company called Protect AI. It's awesome. Everyone should check it out. And what we heard recently anecdotally is a very large financial institution which you all should know the name of, is using five vendors in this space, each for a different use case. They're using protect AI for scanning model code, which is scanning malicious code in the model and open source vulnerabilities. They're using a different vendor to detect hallucination. They're using a third vendor for data leakage and privacy and another vendor for prompt injection protection and a fifth one for something else. What this is telling us is even large financial institutions, immature customers are still having a tough time deciding which company is a long term vendor to bet on. Because all of these startups are at series A or series B stage and they're placing multiple bets in the hope that one of these startup will break out and take a platform approach and absorb the other use cases which I'm really hoping is protect AI. And what can founders really do to acquire real ARR and not just experimental revenue? Let me skip a few slides ahead. Well, the two obvious buckets, one is product and one's go to market in a product category. I encourage founders to think about ways founders and operators, I encourage you all to think about ways that do not rely on the foundational models and the infrastructure that's progressing so quickly underneath. We should look at ways approaches that go beyond that. And one really good approach is really capturing the shipping velocity of product and features that your customers really want. Take together AI for example. It's an AI cloud company that serves an open source community. So every time Meta releases a new version of Llama, it takes the least amount of time to get that model ready and able to deploy on its inference APIs and GPU clusters. And the last round of release there were actually a release partner with Meta. And this type of velocity can really capture customer love. A second approach can be just really capturing the unique data that others your competitors won't be able to capture. Legal tech is a really good example of this. Even up which is in the personal injury space of legal tech has been able to accumulate very disparate information about settlements in the personal injury space so that lawyers who are serving their clients have the best data set to find the best settlement for their clients that a lot of other personal injury lawyers won't be able to achieve because they don't have access to this type of data. And lastly another approach, if you're in the vertical, if you're vertically focused on specific industry or user Persona you can develop a workflow that can be so entrenched that targets this user Persona or industry so well that users can't live without it. Take Runway for example, an AI video generation company. Its market is actually quite broad. It serves advertising, marketing, media, entertainment, Hollywood. And take Hollywood for example. It has such a specific unique set of process, set of steps when it creates a video. And Runway has taken a very broad product and generated really specific workflows for its user base on the go to market side, which I won't go into too much detail because that's Rajen's area of expertise. But a few things to point out here. Everyone probably recognizes hugging face or LangChain and the reason why you do that is because they've done the top two approaches so well. They've married the success of building an open source user community and the network effect. The network effect of having that community. A name that you probably did not expect to see on here is Palantir. Not sure if anyone has heard about Palantir's recent success stories, but it's unlocked a whole new different go to market motion. It came up with this idea of AI boot camps because they're trying to solve the last mile problem of idea to use case and the whole idea of the boot camp is within five days of interactive on site sessions. They will take their customers, their clients, where they're actively engaged with from idea to a real use case that can be deployed. With that I'll turn it over to Rajan.

Thank you Emily. So how do founders, AI founders in particular build a moat and how do they become more interesting and make their companies more interesting for investors like Emily to be considering them? Especially when the product modes are short lived and very affirmable, they have to look behind, they have to look beyond the product mode. And the biggest mode that AI companies are leaning on right now is that go to market differentiation and go to market mode. Raise your hands if you heard someone mentioning the word AI agents or someone who know who's building with AI agents. Like quite a quite a lot of you, almost everyone. And that's true because we are seeing right now is every company is now transitioning towards becoming an AI agent company. This market map didn't even exist six months or one year before and even bigger enterprises or incumbents are now starting to enter into this space. If you follow the news just last week Mark Benioff from Salesforce CEO said that they're making a hard pivot on AI agents and this is not the only category. If you see the entire AI infrastructure stack has exploded in the last one year. Every layer or every category has seen more and more competitors coming in. One particular example is vector databases. Vector databases like Pinecone, Viviate, Zillow and Chroma and many others came up in the last couple of years as the AI LLM models become more powerful. We also saw a lot of incumbents like Oracle, Redis, MongoDB entering the same space because they're already a database provider and now they started offering the AI vector services and the space has become really crowded and we would expect that something like foundational model, where there is huge compute and resource intensive needs, we would see less disruption. But what if you have seen is that every technology company is right now one upping each other with the biggest model and the best model out there. So how do you win in this market when the technology innovations are disrupting every every week and there are new products coming in every month? You have to think about creating a moat. How do you build a moat? How do you think about winning this category? And how do you think about winning more market share as quickly as possible when the market is so frothy? In this session I'm going to talk about four different go to market strategies that we have helped our portfolio companies which is applicable to every AI company whether you are a 0 million, 10 million or 100 million ARR. The number one thing that has worked for us is how do you create a movement and then use that as a differentiator and then create an entire market perception around it. This is not your typical product positioning SWOT analysis. As a founder you need to have a conviction on why your customers think that you are going to win in this market in the long run and then using that conviction to create a movement around it. And one example that has worked for us is at Cohere. When I was leading marketing, we had big competitors who were raising billions of dollars. They were going after consumer bots, chatbots, anthropic was going after AI safety. At Cohere, our belief was that every enterprise is going to transform that AI strategy. So how do you become relevant for this audience? How do you create a movement around it? One of the things that we realize is that enterprises want very different things than what ChatGPT or when some of the AI safety research on more theoretical approach had. Enterprises want security safety, they don't want to send the data or APIs and they want to make sure that their data remains in the firewall and solve their real business use Cases versus a chatbot that will help you write poems or do your homework. So that led to our entire go to market strategy which we feed it on to the product team to build the product for us to have that differentiation in the market. So think about it. Our go to market strategy was leading our product roadmap, which is very different than what you see in the SaaS market, in the B2B SaaS market. And one of the things that allowed us to do is that it established us as the enterprise foundational model and differentiate ourselves in the market. Even though we had some of the bigger competitors in the market who are raising tens of billions of dollars. Now when you have created this movement, how do you change the market perception with it? I asked founders one question. What is that one headline that you want to see in New York Times or Wall Street Journal? Then how do you think about taking your entire marketing strategy and go to market strategy to achieve that? Through series of manufactured as well as series of different models moments that you have on the marketing side At Cohere, we use our differentiation, our belief and our go to market and our movement to get that out in a very methodical way, going after every channel that we had. And because of that, we were able to differentiate in the market and say that we are the enterprise foundational LLM model. The second strategy that has worked for us is, as I said, product modes are ephemeral. You have to build a brand beyond product. And how do you do that? There are two ways. The two examples that I want to share that will be very relevant for every AI companies out there. 1. At Cohere we wanted to be the enterprise player. We were seeing a lot more of our competitors building our consumer chatbot. And how do you gain market share in the chatbot business for the target audience that you you want to go after, Particularly for enterprise, we want it to be the enterprise chatbot. So as our product was being built, one of the things that we realized is that how do you capitalize on this movement and be part of this conversation that's already happening in the market? We pre announced our enterprise chatbot and showed demos around how this would solve real business use case with the proprietary data that you have in your system. This led us like huge amount of lead gen. For us, one of the highest lead gen driver was this pre announcement and helped us get conversations into bigger enterprises from Fortune 100 that we couldn't have before. Not only that, it also allowed us to capture a huge amount of market share that drove enormous amount of conversations that are happening in the market and feeded us into that. So one of the things that in this market, what companies are going to win is that you have to have like a clear differentiation. You have to create a movement, then you then create change the market perception and then getting your share of voice out through some of these tactics. Another tactic that has worked really well was how do you now build a brand around your CEO or your superstars? Every company would have superstars. How do you now take them and build a brand around it? I remember when I started at cohere, our CEO used to give talks in front of a small 20 person hackathon teams. One of the things that we realized is that there's a huge value that we need to create a brand systematically and methodically over a period of time. We invested in making sure that we are starting to pitch with every reporter out there, with every enterprise gen AI event out there. And we build an entire calendar. At one point I remember that we had we were doing around 30 to 40 speaking sessions a month and our entire executive bench was flying around the US to speak at these sessions in front of our target audience. And this takes time. It doesn't get built over a period of one day or two day. It's a snowball effect. You have to start somewhere. And I remember just late last year Cohere CEO was speaking at DAOS World Economic Forum which is where every decision, decision makers are there. AWS booked an entire stadium during re invent for him. So it takes time to build that brand. But you have to invest in a systematic and a methodical way to achieve that. And the best part of it is I spent zero dollars on brand marketing. The next part is it's very important to not go sequential in this frothy market. You have to capture the market share. How do you focus on all the three go to market motions simultaneously in order for you to blitz the market and make sure that your brand is out there and you're getting the adoption that you need. And at kosher and at together AI we are doing the same thing. We are starting the bottom sub motion. It's very complementary to our tops down enterprise sales motion. And we also started with our partner motion. At the same time you started creating awareness in the movement and then building credibility and the brand around it. All these things needs to happen simultaneously. Previously in B2B SaaS, if you remember, we started with one layer. I remember my days at Heroku. We started product LED growth completely and then we scaled it and Then after a few years when we reached around 25 to 30 million, we layered on our, we hired our first sales rep. Similarly, like There are successful B2B SaaS companies who have done this and who have started that one motion and scaled after several years and added another layer. For example, MongoDB did the other way around. They started with Enterprise Motion and then they layer on the product led growth motion. Now this is not the luxury that AI companies have today. They have to go all the three motions simultaneously and they have to work in a complementary way. For example, at together AI and Cohere, our bottoms up and tops down motion work extremely complementary to each other. We had our bot, we had, we had intros to CEOs through a tops down motion. And in the first or the second call they would invite their AI engineer or their AI head of AI and ML. And that's the technical audience that if they're already adopted your platform, you would see that you, you would be able to close the deals way faster. One other big mode that for AI infrastructure and application companies that I've seen is AI is a game for partnerships. Everyone is looking to capture more market share. How do you leverage your technical capabilities or the things that you are good at and partner with several strategic as well as other partners. There are two types of partners that we lean into. One was revenue share. How do you capture revenue share and enter into the customer base. And the other one was completely market share. How do you just penetrate, get in front of your audience and create that moment? All these three motions have to work simultaneously and you have to invest in that at the end of the day. The last one is unusual growth levers. One of the things that I realized is in AI you cannot rely just on your growth teams to find unusual levers for you. You have to have some outsized impacts and big bets that will allow you to capture the market and drive the adoption that you need. I'll give you two examples. One is it's not related to the product at all. Everyone remembers the AI doomsday narrative where several luminaries in the AI space signed a bill saying that AI is going to be catastrophic to everything that we are doing here at Cohere. We believe very differently. We wanted to make sure that we are focused on the most immediate problems that enterprises and businesses are solving. So one campaign we ran was how do you counter this view in the market? Not just by responding to it, but by influencing the decisions around it. So there are a few things we did. We made sure that we are writing opinions around it from the most prominent people in our company, getting that out in the market at the same time inserting ourselves into the policy discussions that are happening in White House. Policy discussions that are happening. Canadian AI Esco here is a Canadian company and change the narrative around it versus just reacting to it. One of the things that we always forget is that how important it is to be part of the solution versus trying to react to that. And there is one campaign that we ran which drove enormous amount of awareness for us in the market and drove market share for us. If you see this was way bigger than any product that we any product launches that we did and that drove us enormous amount of coverage as well as enormous amount of conversations because we were solving real problems for enterprises which was our target audience. The second one is marketing leaders are now going to need a real big long term bets and they have to make long term bets. They need to understand where the market is going. They need to understand what are some of the strategic bets that they want to make and then execute on that one bet. Even this was even before ChatGPT. One thing that I realized was there's going to be a shortage of skilled AI developers. How do you create a gen AI economy? By training existing developers to become AI developers and create that movement around it. So we hired educators, we hired our developer relations team to build a course which is LLM University. After a few months it was the highest driving traffic for us. Drove more than 40% of our entire traffic on our website as well as driven leads. Not only that, I was on a call with one of the CTOs for a prominent stock exchange and what he said is I want to try your product and I know my team is already learning using LLM University so they would be implemented very quickly. So these things are complementary. You'll get more adoption, but you'll also get a lot more in terms of accelerating your deals with that. These are just the four strategies that have worked for us and our portfolio companies. I know we have thrown a lot at you at this and we wanted to make sure that if there are five things that you want to take home with you, these are the five things you need to go through.

We learned today that this new wave of Genai startups have a different type of growth profile and the big part of that, or at least some part of that is due to the fact that of experimental revenue and you can tell there are different signs that point you to what experimental revenue is in retention, in sales, motion, in usage and Use cases and founders. In this new era of gen AI, you think about new ways to build modes and differentiation for both products and go to market. But we're arguing that go to market is becoming more and more important.

And the next generation of AI companies, they're not going to be product led, they're not going to be sales led, they're going to be marketing led. So every marketer in this room should feel really proud that you're going to be part of this big movement and the role of the CMO is going to evolve and they're going to be more strategic, they're going to be shaping the vision for the company, they're shaping the go to market plan as well as the product vision for the company. And with that, thank you so much for attending and we would be. We have a few minutes. We will be able to take questions if you have any.

Thank you so much.

Thank you so much.

We have mics at the front. If anyone's questions.

There'S a mic there.

So I'm wondering how for Genai folks.

Are using pilot programs towards scaling and how you see that opportunity where you.

Show the business the value of a.

Genai use case that leads towards conversion.

And is there a workflow or a.

Methodology that works differently than traditional pilotable.

Projects in the past?

I can take that from the go to market side. I love Emilia's opinion on the investing side, so it's a really good question. One of the things that we realized is that sometimes, in fact, a lot of the times the customers don't even understand their use cases well or they don't even understand what they can solve. With Genai, one thing that we did was the pilot program was that we used to do a lot more PoCs with our enterprise customers and other customers. Customers. We realized that it was taking them a lot of the time and the cycle was getting just prolonged. One thing that we did was we started charging for POCs, even if it is from creating urgency, just so that they are able to invest their time and work with us. Because a lot of the times when they say that, oh, this is something I'm exploring now, when you make the POC charge, they're actually more invested and we give credit. Once they close the deal, we would refund that credit but it's like of charge credit so that they create more urgency around it. So this piloting but at the same time with an intent to convert more faster.

And I also really want to go back to a previous point I made in terms of making sure. That the customer you're doing a pilot POC with has a real production use case for your product. A lot of customers are just trying things out sometimes because they don't. They're not ready to go into production yet, which is totally fine. But for the ones that you dedicating a lot of your team's effort, make sure it's a live production, that your product will be used, which means that it's a lot harder to rip out and you'll likely have expansion opportunities down the line. You have one question.

Thanks, Chandran, Chief product officer of PapFactory. And not only are we a Cohere fan, but I think there's one question, sort of the elephant in the room, which is all of the strategies that you displayed were great when you have.

A lot of funding.

Many of the folks here, and especially the way we're working, is kind of a cost efficient way to get started.

Especially even if you're an AI LLM company.

So what is one strategy that you.

Think is probably the most efficient, cost.

Efficient way to get your message across?

That's a great point. Right. I think there's an enormous amount of funding going on in the AI space. One counter argument to that is if you compare Cohere with anthropic and other OpenAI and others, you would argue that we got very little, we raised very little money and most of the funding that we got went to compute. So we were operating on an actual very lean budget compared to some of our competitors out there. So there are a few things that have worked really well for us. I think we have to find unusual levers, whether this is creating a brand out of the superstars that you have or whether you're using partners to distribute overall in terms of getting more adoption. Because a lot of the times if you are good at something, your partners are willing to invest more resources because at the end of the day they want more complementary solution and customers on their space. Those are a couple of things which are very low budget. It takes a little bit of time to execute on that, but it will not require a lot more money for mbu.

And being an investor in Cohere, I just want to mention that one thing that really surprised me was Cohere really leaned in on the Canada narrative. Cohere is kind of the brainchild of AI in Canada and really leaned in on that. So the government and everyone in a Canadian ecosystem, from a talent perspective, are kind of supporting cohering away the anthropic OpenAI and a lot of the other model apps in the US are not getting.

Thank you for your talk.

My name is founder and CEO of Emotions, an AI coaching company to help sales teams members with regards to reduce ramp up time.

Curious your statement, the next generation of.

Good AI companies marketing driven instead of product or sales driven. Like why do you think that's the case and how do we make the.

Marketing, how can we do well in.

Marketing to make that a successful company?

Yeah, that's a great question and I think there are two aspects to it. One is how marketing and sales are leveraging the newer technologies to be more efficient and drive more momentum and adoption with very less resources. The second aspect to that is also how do you how can marketing and sales be more strategic? I think the typical playbook is very well defined and what I found is that those doesn't work in AI startups. One of the things that has really worked for us at Coherent Together AI is that I am guiding the product team on. These are the product moments you want to launch. Now how do you build and what narrative should I be going out with from your vision that you have? So we systematically launched certain products that the product team would have never thought that you can package these things together and launch it. And our sales team was super happy because, oh wow, now it actually solves a customer problem versus now I'm just selling features to them that are more incredible incremental.

I think we might have time for two more questions, so I'll go over here first.

Okay, thank you very much for your talk.

Actually my question is for you. What's your vision from an investor point of view about this experimental revenue but from SMV perspective.

Right.

So we're in that space and we're.

Seeing a lot of interest, a lot.

Of inbound leads, we're saving the cost, customers, they're getting the value very, very fast. But we are always hearing like watch out for the churn because of this same kind of concept.

Right.

So your question is from an SMBs perspective, how do you tackle those. Those issues?

Yeah, exactly.

Like more like we are targeting SMBs and I would like to understand your perspective about experimental revenue in this space, not just in enterprise.

Right?

Yeah, I think SMB, it's tough because by nature of this market segment there is likely more churn. I think going back to the point if you remember of the company with the two different products and one of the product that is doing better in terms of trending on the Churn side is that they're qualifying their leads really, really well and That I can say that. But what does that actually mean? I think to the point of a few different ways of getting to that answer of like exactly what is your customer looking for and what is the product that where is the product resonating with the most in your SMB audience and you're trying to look for real production use case like I'm saying again. And I think SMBs are probably more likely to try out a few different products before they commit to one. So I think really capturing them at that inflection point in terms of when they're making that decision, which is a long term partner that we're going with I think is really key.

We have time for one more question.

Hi, my name is Sagar, founder and.

CEO of Host Ready.

We provide virtual host for small businesses and big brands.

So these agents not browser based or.

The phone lines imagine like a call center for restaurants. So we are partnering with Square and companies like that and yeah, we want.

To spread the word out. So what is your criteria for this? Hyper growth? Yeah, so at Hypergrowth we partner with all companies. We have early stage, mid stage and late stage and we help them. So we have an AI accelerator that we run. We have a 2024 cohort going on right now and we will start the 2025 cohort pretty soon. So we help AI companies, we partner with founders for early stage companies as well as some of the late stage companies with their CMO and CRO and help them on finding out the right strategies on the go to market side, whether it is on marketing growth or on the sales side of it. So we have an AI accelerator for that that we help early to mid stage AI. So your focus is on the gtm? Focus is on the gtm.

Yes. Fantastic. Thank you.

Thank you so much. I know we are over time.

We'll take questions after the presentation.

So that's good.

Thank you so much. Really appreciate it.

Thank you so much.

Thank you for attending.

Sam.

It.

Sa. Sa.