LIVE SaaStr AI Day: How Marketing Can Find PMF and Growth with HyperGrowth Partners

A quick intro and then I'll let you really run it and then I'll help at the end for any Q and A. But yeah, thanks again for closing it out.

Sounds great.

Okay, awesome. All right. Hello everyone. Welcome to the last Saster AI Day session of the day. You've made it through the whole day. I know some of you have been on all the sessions. I think we had seven or eight jam packed sessions back to back. So thanks for joining us. If you missed any of them or if you want to watch them later. Just as a quick reminder, we are live streaming so it's on the Saster YouTube channel. All the sessions will be going up on Saster.com shortly. We'll just edit them, get them up on Sasser.com so we're excited to have a strong closer for you guys on marketing and product market fit. Rajan, I know you've been to Saster. You just did Saster Annual last year. Now you're doing Sasser AI Day. Thanks for doing this. I will defer to you since you're the expert here, but thanks for joining us and I'll let you take us home.

Thank you Amalia and thanks for having me here. I'm super excited to be here and thanks a lot for waiting for the last session and quite a big crowd. I'll introduce myself. I'm Rajen. I'm the general partner at Hypergrowth Partners where I run the AI go to market accelerator. AI the term that we're using or hearing the whole day. And at the accelerator we help genai companies from early stage to growth stage on that entire go to market marketing growth and sales. We have a portfolio of 40 companies from top companies like Together AI, Deepgram, Composio and few application companies. But overall we have companies like Honeycomb, Vercel, Ramp, all of that in our portfolio that we help advise. I currently serve as an interim CMO for Together AI and AI infrastructure company and have led marketing at Cohere, one of the foundational model companies. And before that I ran marketing, growth and the self service business at Heroku, a developer platform. You'll see a lot of examples from some of these companies, but I'm excited to be here and let's get started. So how to find product market fit and growth for AI companies? We'll cover a few strategies but before that let me walk you through what a typical cycle used to look like for a B2B company. A sales cycle. So very linear, very sequential. You build a product, you find product market fit, then you continue to grow and scale and at some point in the maturity and expansion you add new products. And marketing always used to be a lagging function when you find PA product market fit. And while this was true for traditional B2B SaaS company, it is no longer true for an AI company. You'll see multiple of these cycles just compressed into their entire life cycle. And that's because there's technological innovations, there is new market entrance and the whole landscape is changing very fast. And you don't have a luxury if you are an AI company to follow the sequential path and marketing and go to market in particular would be very early in helping finding product market fit at every stage of this cycle. And that's basically what I've seen at every company that I've advised at. So I saw a lot of these, a lot of the sessions. Early on everyone talked about Deep seq. So let me just make sure who has not heard about Deep seq? Raise your hands. Almost no one. Right? Everyone has heard about Deep SEQ this week and what you can see is like the technological innovation that is abroad is disrupting all the existing incumbents in the market. And this is not just happening now. It's been happening for the last few years since ChatGPT. Like we saw open source models come up, we saw more closed source models come up, we saw like the gap between open source and closed source started reducing and now we have more open source models that are actually beating a closed source model. So the trend has been like there's technological innovations is helping drive new market entrances, helping drive more competition and challenging all the incumbents. It's happening across, right? Not just foundational model space, every company's becoming an AI agent company. Like this market map didn't exist like 1 1/2 year ago and now every company, new and AI native companies as well as like bigger companies like Salesforce, all of them are becoming an agent company and there's enormous competition going up at every layer of the stack. And very similar like this Chennai infrastructure stack has become so crowded now. How do you think about like if you're starting a new company or if you're an existing company pivoting into AI? How do you think you can emerge from this noise and become a leader in your category? What strategies you should invest in to build towards product market fit and conquer the market share and how do you stand out in this market? Right. And there are a few strategies that we'll go through today. By no means these are exhaustive but what has worked for me and what has worked for the companies that I've advised, Number one, creating a movement. So it's talked about a lot, right? Whenever every new company comes up, like hey, I want to find my category, I want to create a new category, I want to figure out where I stand with the competitors. And this is a typical B2B SaaS way of positioning and category creation, right? There'll be some random access that people will create. You'll try to fit your company into the top right section and it's like, oh yeah, we are winning the market. But this no longer holds true. Some random attributes will no longer be relevant because the product is evolving. And these random attributes would be very affirmal. Like within a couple of months you'll lose that position. So what makes you stand out? The number one thing is creating a movement, right? And it's creating a movement behind your vision and how you would differentiate in the market long run. What do you stand for? I remember when I was at Qaheer, we were one of the early closed source, foundational model company. And then we had bigger incumbents like OpenAI and Anthropic coming in, raising billions of dollars and everyone was going after slightly different set of market like OpenAI started with ChatGPT focused on research, going after more of a consumer market. Initially Anthropic was focused on going after the safety and the research side of the folks. So here we had to figure out how do we differentiate in the market. And as ChatGPT came out, we realized that there's more and more need to stand up for our vision which is going after the enterprises who are actually going to be commercializing this technology. So one of the things we did was we created an entire new positioning and movement around enterprise AI and enterprise LLM. Enterprises would need very different things than a consumer would need. They care about security, privacy, more data opt out clauses, and they want to power hundreds of use cases. So how do you think about solving for the enterprise audience versus going and competing head on versus OpenAI anthropic. And that helped us create this new category of enterprise LLM or enterprise AI. And that's what I think every company needs to do in terms of thinking about what is their long term vision and hanging onto that versus trying to come up with a positioning with the current product or with the current technology that they have. Similarly, when I joined together AI, they were going after a very simple research driven artificial intelligence company and they were just building newer technologies that was driving at the systems and the AI research space and Commercializing that technology. Now we had two big visions. We wanted to become the default platform for open source and we wanted to compete with the bigger cloud providers. So we created two movements. One was how do you create a movement around open source building with open source models? And then how do we pitch together as the default platform to build with AI? The other movement that we started creating was how do we become and compete with the cloud providers versus like some of the point solutions in the space. So we thought about what would represent us and what category do we want to go after. And we wanted to create a movement around accelerating AI for AI pioneers. And then we coined the AI acceleration cloud which is very different in full stack of helping every companies to build and grow their entire generative AI lifecycle, whether you want to train models, whether they want to build innovative applications, or whether they want to just fine tune an existing open source models. So both this movement worked hand in hand. The summary for this is you can create a positioning, but it would not be very relevant if you don't have a movement around it. And once you have a movement that you want to anchor around, the next thing that you need to do is now how do you use this movement to change the market perception? Because the market perception matters a lot in this space right now. And how do you punch above your class? So one thing that we did at Cohere was we used our movement on enterprise LLM and we used all the different tools available for us. From the marketing side, we went after every reporter to go and pitch our enterprise positioning. At one point I remember we had 30 different speaking panels in a month and everyone from an executive bench to the directors and even engineering team were speaking at different AI events for enterprises that were being hosted. And all of that came at like $0 to the company cost. Right? And that's how you try to change the market perception. By systematically going after different areas from the marketing channels and then conveying that new movement that you have just created. The second thing is now how do you build credibility beyond the product? This is actually really important because as I mentioned before, product modes in AI are ephemeral. It's hard to sustain a technology advantage beyond three months. It's very important to build credibility that would drive beyond that few product features or product launches. And using the marketing vision, we have to think about how do you promote not just what, what's on the truck today. One of the examples that I can share is as at Cohere we were going after more of an enterprise audience. We realized that OpenAI was going after ChatGPT and consumer market anthropic launchcloud which was more focused on consumer at that point. How do we think about building and creating a space for us as an enterprise chatbot and knowledge assistant for enterprises? Right. So one of the things that we did was we packaged some of the features that we had and we announced our enterprise chat visions month before the product was actually ready. And this became like the biggest lead gen driver for enterprises that we were targeting because there was nothing in the market that would allow them to leverage their existing data sources within their proprietary systems and build a knowledge assistant for them. And this became like one of the, not only just one of the biggest Legion driver, but it got our huge share of voice in the market and it drove really good word of mouth for us, especially for the enterprises that we cared about. And they became like our early beta testers and they became like our enterprise customers overall. That's like one way of doing it, right? What's another way of doing it? You need to build a brand around your CEO and superstars because again your products are ephemeral it tomorrow you'll have no advantage about your product, but you will have an advantage about the brand that you build with the people who are building this technology. So one of the big things that we did at cohere and together AI was how do you identify the top researchers, the CEO and build a brand for them across different channels. I remember like when I joined Cohere, Our CEO was speaking this was back in just before ChatGPT was launched or CEO speaking at smaller developer community events with 20 developers and over a period of time as we built his brand last year, 2023, the year before now 2023 AWS re invent AWS booked an entire stadium for him as part of that. And it doesn't happen overnight, but you have to systematically start investing in the people and building their brands. And similarly with together AI, we have amazing researchers the CEO and how do you drive that vision and how do you drive that brand and scale that because that's going to be the long term mode versus the product that we could not launch. That's another way to do that. Third, simultaneous go to market motion. So I worked at several B2B SaaS companies and every time I've seen like it's a sequential thing. Like you start with one motion, whether it is the product led growth motion, bottom up motion and then you layer on some of the enterprise motion or other motions at Heroku when I was there. We added our first sales at like when we reached like 25, 30 million. ARR and database companies like MongoDB started enterprise and slowly scaled their bottom up motion. And a lot of these other companies that you see here started had their like false sales and added top down layer like few years after they were like well into product market fit. What's happening now is it's no longer a luxury. Like you don't have that much time because the product cycles are so fast. There are so many different competitors coming up so systematically. You have to now start investing in all the go to market motions together. You have to build your awareness, you have to drive your moments, you have to build credibility and brand for your CEOs, for your top researchers and even for the top people in your company. But now you have to have all the three pillars of your go to market motion. You have to start your bottom of product led growth motion, going after users, going after the tops down sales motion. And then if you are in the infrastructure or a company where you integrate partner and channel motion becomes really important. And at together AI at Cohere and other companies that I partnered with, all of these three motions came in handy. And I'll give you an example at the bottom of motion that we started initially at together AI. Now we get more monthly signups than we had all time signups the years before. So just in the matter of couple of quarters or few quarters, we are able to scale our bottom up motion. Going after developers, going after using our rel motion, building open source applications because that's the audience that we want to to go after. And at the same time we layered on some of the top down account based marketing both from inbound and outbound side of it. Like we ran campaigns around going after targeted accounts at Cohere very early. And similarly we are starting that Now@together AI. And some of the early customers that we got was through some of our account based marketing and top down motion. And that's how we were driving more product market fit early on to get those fast lighthouse accounts, to get those fast inbound customers with more marketing and more reaching out. The third is like partner in the sales partner channel. This is over the period of last couple of years has become extremely important. Like everyone in the AI space is starting to now compete with you, but at the same time they want to partner with every other company across the stack. For example, all the cloud providers have the same capabilities, they're going after the same market, but they still want to partner with with Every other infrastructure provider, every other foundational model provider and you have to take advantage of that. How can you think about the partners above the stack and below the stack and leverage them as a mode And I've broken down like few different categories that has helped cohere and together AI, there are a couple of partners who are very strategic. Like you have to have partnership with them, especially when they are below the stack or above the stack. There are cloud providers, there are platforms that will allow you to distribute your technology to their own users. Then there are open source platforms like LangChain vector database which will complete the entire solution of the use case that the customers wants to build. For example, if you want to build a rag application that we have saw in a few other sessions before, you have to use multiple of this technology, you have to use a vector database, you have to use a foundational model and you have to integrate all of that together with different embedding models, whether it's the re ranker models and then you have to integrate with your own data. So partnering with different ecosystem partners become extremely important in this space. These are the three different go to market motions that has helped us and all of that we started together. And one of the other things that I really found valuable, especially going after the bottom up, top down and the partner channel is all of these, especially for AI companies are all complementary to each other. So a lot of the times we think about whether there will be a conflict, whether there will be cannibalization of my sales if I go through partner channel. These are all complimentary and you have to invest in it early on, otherwise you're always going to be lagging in terms of the new growth areas for finding product market fit. The last one, I think this is very innovative in the sense that I think we had this term growth hacks five, eight years ago. It's drawn into more of a playbook for B2B SaaS. But at the same time I think what has changed with the AI space is that these playbooks, while helpful, help you get the growth. You have to have more long term strategy built into some of these growth levers now. For example, you have to think about that your target audience is broader than your prospect list and you have to think about how are you going to reach that and how they care about you more than just your product. For example, if you remember like a year ago there was a huge series around AI Doomsday and everyone was trying to lean into how AI is going to be very detrimental to the humanity And I think a lot of that, while it's a great narrative out there, like now, how do you think about how does it relate to your product and your company and your vision? And at Cohere we were going after the enterprise and enterprise market. No one was actually solving the challenges of the enterprise companies or the enterprise products, which is bias, which is misinformation. So now how do you provide a counter view to the market that while some of these long term concerns may be true, we are nowhere near that. But what people should care about and what your audience would care about more is what are some of the near term risks that would affect their customers. That angle drove huge amount of share of voice for us at Cohere. And we systematically went after build those partnerships with other enterprise companies. And we also figured out a way to then join White House for AI commitments for enterprises. So it becomes like a very, very strategic move in the sense that what do you care about as a company and your prospects are more than just your target audience, is more than just your immediate prospects who are becoming their customers. Similarly, recently, a few months ago, we had this new SB 1047 bill in California. Now how do you think about not just being part of the conversation but also shaping the conversation, which is the more of the way to say that you cannot be a passive reader or you cannot be a passive listener and react to these. I think you have to shape these conversations that are happening to find that initial leadership position in the market. So all of that, right? We did a lot of that at Cohere, we did a lot of that at Together AI. It not just drives your awareness, but it drives a lot of word of mouth and share of voice overall for us. And it drove more than any other product launches that we had. And that's how systematically you try to find out what's next and then you make some long term bets and then you execute on those long term bets. Even today, before this call, we were setting up calls and broadcasts for our CEO at Together AI to comment on the deep SEQ open source models because that's like so integral. Open source is so integral to the part of what we do and what our vision is. And it was very important for us to act on it and be a voice versus just be a reaction to what's happening to the market. The last thing that I wanted to share is how do you make more long term bets, right? And as marketing leaders, a lot of the times it's more than just executing on some of the company visions like for example, when I joined Cohere, big thing was for us was the technology was so new. This was before chatgpt times. No one knew how to what an LLM is back then. It's so weird now, you'll not even imagine that that's happening. But that was true at that point. And even if we had a great product, there were no one in these companies to actually implement this technology. So one of the big bets that we took from the marketing side was we wanted to build something which is called an LLM University now. And we systematically build a lot of content around it and we packaged it as a LLM university. This was before ChatGPT times. And even after the hype, the conviction about the skilled resource gap in the AI market become more prominent. And over a period of time that generated 40% of all of traffic. I was in a call with one of the CTOs of a prominent stock exchange in the US and he was bought in from a tops down enterprise perspective. And he wanted his team to ramp up and he wanted his team to start using the product to experiment with it. And he was like on the calls, oh, I see you already have LLM University. Let me just send it to my team to figure it out. So this is what I say, like long term bets work systematically. Even if you're going after like bottom up as well as the top down, go to market motion to find that initial product market fit, but allows your audience to start using your product. So these are like just four strategies that I shared and there are so many more. Right. And the few takeaways that I have found which is helpful is in this new era we need to think of new ways to build modes and differentiation and go to market is going to lead the charge here. And the next set of successful AI companies will be marketing led because everything like your product, your technology and the entire markets are in constant flux. Your vision is changing every day with new things coming up. And what I recommend as you're thinking about someone to come in and lead your marketing, that your CMO or your head of marketing role becomes so crucial because a lot of the times it's starting to drive company vision and starting to place few long term bets early on and to drive this sustainable long term advantage. So the key aspect is like the role of CMO has changed. It is more than just going after the marketing, but it's more becoming strategic. And there's an intersection between a CMO and a CEO role right now that I see which is becoming more prominent to Drive real growth in this high growth AI companies. With that we are open for questions.

Yeah, thanks for doing this. You've got some good comments here in the chat and then question from the live stream is, are there any AI tools? I guess the question is, sorry, what are you using in your marketing stack or any tools that you recommend? Kind of a two parter, but yeah, yeah.

So a lot of the marketing stack that we have built at the companies that I've been part of is to drive more efficiencies. For example, everyone now started have started to use more copywriting tools to help them scale content. We have started to use a lot more on the prospecting side, especially when you want to scale your outbound, you want to scale your enterprise side of it and integrating that within your stack becomes really, really important. Other things we have started to use are more focused on driving more insights from the market. So we are starting to use technologies to scrape the news. Tell us immediately what's happening in the market. I'll give you an example. The compute market is changing. The new chips are coming every day. So now how do you think about who's actually using these chips? So we actually scrape the hugging face repos for what models they are using, what models people are training using some of these tools and then we use that and feed into our outbound motion. So a lot of the times the early use of the tools start from an efficiency perspective, from whether it is from a lead gen perspective or from a research perspective, but extending into a lot of helping accelerate the things that a product marketer or a growth marketer would be doing. Yep.

Any, any tools in particular that you'd like either previously or now as a follow up question and then we've got some more questions.

Yeah, definitely. So some of the tools that I found really helpful especially on like the prospecting and the outbound side are clay.com, unify as well as Koala. So, so those are like really great tools to help you like really drive more efficient go to market process.

Cool. Let me bring up Vishal to ask his question next. Vishal, do you want to unmute and ask your question?

Yeah. Amelia, thanks for, thanks for that. My question to you Rajan is how does this translate into a positioning exercise when it comes to specifically the buyer, the AI buyer, enterprise buyer being flooded with so many different tools that are out there. Right. So how do you even educate them even positioning? Because they need to have some evaluation criteria based on which they might measure or compare products. How does that work from An AI context. So I think the typical product positioning or the differentiation that we are used to, how do you compete with or how do you compare with certain competitors? I think that would not work. I think the way I would position that is what is the long term vision that you want to go after? So if you are a specific application company or a specific infrastructure company, what is your envision for together AI? The envision for us was to compete with some of the cloud providers as the generative cloud. Generative AI cloud. So now the way you think about it is how do you create that, create that movement, Right? So from the positioning perspective, our vision is this. Now how do you get to that vision? What would customers care about along the journey that you have a hypothesis on and then you list down all the different things that they would face as problems. I didn't even look at some of the competitors while creating this positioning. For us, the vision was we want to accelerate all the AI workloads for all developers and enterprises. So with now, how do you start thinking about what you have today, but what you're going to have in the next 12 months or 24 months as a product side of it? And using that and just translating that into a narrative helped us define the entire positioning for us. So I had big sessions or long working sessions with the CEO, with our research team, with the founders. Where the technology is going to go, what are some of the hypotheses we're going to bake in? And using that and transforming that entire thing into a narrative was super helpful. I didn't use any single line item in my narrative doc around features because those are just non existent at this point because by the time I finish it, it's going to change. So how do you translate that vision into a narrative and then you identify one or two big angles. For us it was open source and for the second one was accelerating the AI deployment using that now you create movement around it. So thinking about the vision versus the features is the first thing that you need to start doing. Gotcha. And if I may ask, how does that differentiate from a more enterprise buyer segment? For example, six figure deals, seven figure deals that are the buyers may or may not be educated. There may or may not be a various competitors or they're misinterpreting your product versus somebody else's product. Any ideas on that? Yeah, I think that is more from the capability. I would say that is still more from a capability standpoint. What are you going to achieve then? Every enterprise deals like six figures, seven Figures, deals that we had would have a roadmap. This is what is it now, but this is what is going to be in the next 12 months because everything from our compute layer, everything from the platform layer to would be drastically different. So how do you incorporate that into your pitch when you're pitching to a specific customer? Becomes like really important.

Thank you.

Cool.

Thanks for the question. We'll go to Alice next. So let me bring her up.

Hi there. Hi. I'm so happy to hear that you feel the time has come for marketing to enter the arena for AI companies. You know, as a marketer myself, I've seen over the last few years all the funding that has been going to startups has been on building products and acquiring compute power. Benioff has been talking about their switch and focus this year to hiring I think 2000 sales and marketing people to push agent force. I'm just curious though how you advise folks to find each other because I think LinkedIn as a model is broken for connecting talent with hirers and founders. And so yeah, I'm just curious about, you know, how do we find each other? It is, it is actually like fine. I think I would say slightly change the problem statement. Like it is difficult to find the right partner on the each of the side of the table a lot of the times. Right. There's a lot of noise around. The number one thing that's helped me is going to some of these meetups and conferences like even like the annual Saster was amazing last year when I met so many different founders and there is a site called Cerebral Valley AI. You would find so many different events, more networking opportunities there. Like I talk to at least three founders every week and it is more from my knowledge perspective like what are they building and how does market is evolving. That has been like a number one source for me to try and find some of these great AI companies that I help with. Great. Awesome. Thanks.

Yeah, thanks for the question. There's a follow up related question on the live stream just talking about CMOS and talent in a slightly different way. Any recommendations for CMOS to stay ahead of the AI curve? It's a big question.

Yeah. I think there are two aspects to it. One is understanding the market and understanding the technology becomes really important right now. I have a technical background. I was a developer in my past life. It drives me more interest reading some of these research papers, understand that and now I just use these research papers to convert it into a podcast and then listen. So that is much easier for me and interpret that But I think understanding the technology, understanding the market would be huge because at the end of the day you have the experience on the marketing side to drive these functions. I think what founders and what some of these investors are now looking for is how can you do that when you don't know where the product is going to be in the next 12 months? And I think if you know the market, if you have built some long term hypotheses on where the market goes and if you, if you have certain belief, like for example, when I left Cohere, my big belief was that open source is going to drive the next generation of innovation. And that's the reason I partnered with together AI and I partnered with some of the other companies because if you have this hypothesis based on your market perception, that drives a lot more conviction and that would drive a lot more confidence when you are talking to some of these companies. Awesome. Cool.

Well, thanks for answering all the questions. I know we went a little bit over because it's the last session of the day. Any parting thoughts on just AI or marketing to leave our audience with today?

No, I think this wave is going to be here. I think either you immerse yourself in this space and whether you're working at an enterprise company who is making this pivot, or whether you are starting to think about working for AI native companies. This is an amazing opportunity for everyone to upscale, but also to be part of this next wave of innovation. And that just drives me nuts that people still are not willing to accept some of that. So please have an open mind and adopt everything that you see. New tools, new technology, read about the new research that's going to keep you ahead.

Awesome. Thanks for that, Rajan. And I'm sure we'll see you back at hanging out, networking, walking the floor, whatever, whatever it is you want to do there this May. So thanks everyone for joining us for Sastra AI Day. This is our last session, but if you missed any of the other ones, you can catch them on the Saster YouTube channel and you'll find all of us back in person at Saster Annual 2025. It's in, as we've been saying all day, it's in May this year. So just, you know, mark your calendars a little bit earlier than the last couple years. We'll have a whole AI stage again, of course, and we're excited to see you all there. Thanks again, Rajan for doing this.

Thank you so much for staying over for the last session.