Frameworks For Driving Sustainable Growth w/ Mike Duboe, Brian Balfour, Shaun Clowes | Greymatter

Hi, folks, I'm Mike DeBeau with Greylock Partners. Welcome to our podcast Gray Matter, where we bring in some of today's top tech entrepreneurs and business leaders to share their story from startup to scale up. Today we're talking with Brian Balfour, founder, CEO and Reforge and former VP growth at HubSpot. Sean Klaus, chief Product Officer at Metromile, former Head of Growth at Atlassian. Great to have you here, guys.

Thanks for having us.

Hi, Mike.

Cool.

So we're talking about the topic of growth and growth teams. It's one that is increasingly common in today's startup world, but it's also pretty ill defined. It can mean something quite a bit different, company to company. So today we're fortunate to have guests who have been instrumental in defining and modeling growth as we know it. So much that they actually teach a program on it with Reforge and could hopefully elucidate many of these concepts for our audience to kick things off. Brian, you've had quite a career transition from founder to becoming VP growth at HubSpot and. And then you founded Reforge, a school that trains the next generation of growth leaders. Tell us a little bit more about your journey going through that.

Yeah, so I started at HubSpot at a pretty interesting time. We were about a year and a half before going public, and I came in to work with a couple other executives on establishing the new products division group, which was all about establishing like new $100 million lines of product to business for HubSpot as they kind of look to the next five years of growth. And so we developed and launched a number of products. The main ones that kind of made it through the pipe was the CRM and HubSpot Sales Pro in that division. I just felt like the startup journey kind of packed into a couple years in terms of revenue and team growth and all those things. And one thing that I started to really realize is I would sit in these one on ones every single week with my team. Somebody would be asking me about professional development. I'd spend hours researching what to recommend them. I couldn't find what to recommend them. I felt like a terrible manager. It was a terrible experience for them. And so through twists and turns, I just ended up deciding I was going to try to dump out all of my thoughts into like a course on the side. I ended up collaborating with Andrew Chen on that first one and we did like a prototype of a prototype, like something super embarrassing, but went way better than expected from there. Decided to like, jump into it. And so, yeah, Reforge is really focused now on unlocking all of this amazing knowledge that's trapped in the heads of amazing practitioners like Sean, like yourself, like the community that we spend a lot of time with and just make it more accessible to, you know, that mid career practitioner who's like really navigating that really important middle part of their career. And the awesome part is that I just get to see across so many different companies of like, what works, what doesn't work. I get to tap into the heads of a broad group of people. And so I just feel like I'm learning every single day.

Awesome. Well, I want to jump more into Reforge later, but just to set the stage. Sean, you know your background is pretty interesting. You were one of the early pioneers in building B2B growth functions. Spinning up what I think was the industry's first B2B growth team, perhaps at Atlassian. What drove you to do this? And I guess as a result, how has your construction of growth and product teams evolved along the way as you've taken on subsequent roles?

Yeah, it's hard to believe, but it was in late 2012, I think, when we started down that road and that's only seven years ago, but it feels like eons. And so I was at Atlassian, I was running one of the major products there, and at that time this whole growth hacking thing was happening. And I really love that terminology and I don't really love kind of what it even stood for at that time, but effectively what was clear was that some people were successfully achieving step change differences in growth by making small changes to product experiences. That's how I internalized it. Other people internalized it as hacking, but I internalized it as the fact that the last mile, like the little things you do in products can drastically affect whether or not they're going to be successful or not. And so that was happening. And I was sitting inside Atlassian, an already very successful company, but by that point Jira, Confluence and other products like that. And I remember Mike Kennen Brooks was like talking about this growth hacking movement and he's like, I wonder if that could apply in the B2B context. And it sounds a little bit crazy, doesn't it, taking this consumer hacking mindset to such already successful products. But he basically said, okay, well if we think those things could apply, then we should try. And so he gave us a six month life to live and set me up to lead this team and basically said, go out and prove it and if it doesn't work, then we'll just kill this team off in six months. If it does work, then we'll see what happens after that. And so I spent the next five years building out the growth team at Atlassian. And obviously it was super successful. And little things we did turned out to drast improve the growth trajectory of an already very successful product company.

Yeah, I guess. What are the conditions under which it would make sense to bring over B2C growth tactics over to B2B side? I mean, I think obviously the long cycle time and actually longer feedback loop in a B2B context could be one of the natural deterrents against building out a growth function. But how do you think about the right conditions?

Yeah, I'd say that it is more often right to do it than it is wrong to do it. And it's just that people don't naturally lean into it. And it's because there are a couple of extra Dynamics in the B2B space that make it seem less likely. So the first one is sample size, so that you tend to have much lower volume. So you're like, how do I use these techniques in much lower volume funnels? The second thing is, do you often have humans in the loop? So you have humans who are in there whose job it is effectively to paper over the ugly parts of your product experience? So I'm thinking of customer success people or pre sales engineering or the salespeople themselves. Right. And so effectively you don't pay attention to these little bits of the product experience that are frictionful because they, the system is set up to make it very hard for you to even see that they exist. And I think that when you step back and you think about it in a different way, then it's just incredibly clear. Once you take away all of those support networks and the safety net that comes from that, that in the B2B context, absolutely the same techniques work. And absolutely, at the end of the day, the person on the other end of your software is a human. Any assumption you have that that person is going to be more tolerant of painful experiences is by definition flawed. And the consumerization of it and everything else like that is 100% coming into full in the enterprise.

I'll double down on that real quick. I think something we see at Reforge is like Everybody in the B2B company feel like they're a special snowflake at the beginning of these programs. They're like, oh wait, no, I'm in B2B. I've got to do this totally different than the consumer people. And then we have to go through this whole process of like, no, like it's still an individual on the other end, you know, where all individuals respond to similar types of like, motivations and emotions and those types of things we're still looking for, like pleasant, frictionless experiences, looking at engagement and retention. And all of these types of things are just as important, if not more important than B2B. So I'm constantly surprised by B2B product or marketing professionals who feel like what they're doing is fundamentally different from the consumer space. We've definitely seen them merge a lot more lately with the bottoms up model and stuff, but that'll continue to blend. But it is very, very similar. And it is not a snowflake situation for sure.

I completely agree. It's like people assume that just because I'm at work I have a higher tolerance for torture. Like, I don't think. I don't come to work and go, yep, now tolerant Sean mode is on and I'm going to be playing around in some big enterprise Y piece of software going, oh, this is amazing. You know, it's still an awful, torturous experience. And so if you think about in the end, bringing a really great product experience that people want to use because it is rewarding is like bringing a gun to a knife fight, right? And it doesn't matter whether or not it's in a consumer experience or in a B2B experience, in the end you should build products that fit their markets. It's that simple.

One of the things that we've discussed is, you know, if growth and everything you're teaching at Reforge and those of us who are running these functions in companies are successful, you know, the growth team as we know it kind of abstracts itself away and it just becomes something that exists within core product teams, core marketing teams, et cetera. Do you agree with that? And if so, people that are maybe thinking about betting their careers on growth right now, how might you think about that going forward?

Yeah, I agree with some of it and disagree with some of it. First is that I very much view growth as a cross functional discipline. It is not a function in the org itself. All growth teams, the successful ones, tend to be a mixture of product engineering, design and marketing skill sets sort of blended into one.

And.

And they more often than not fail when they are established as their own separate function and they kind of get siloed away from certain resources or something of that nature. And what we've seen is like what growth is really about is I think if we Thought a while back, maybe 10 years ago or something, the conversation around growth was very much around just build a great product or people were viewing growth through the lens of like PNLs and things of that nature. And that is not actually how software products grow. There is a different set of things of like how you look at it. And so growth to me is a combination of both looking at growth from different strategic lenses, things like growth loops, which I know we'll talk a little bit about, mixed with a different set of processes or thought tools or problem solving tools. One of the main tools is around experimentation and the scientific method and different variations of those. And so those two things combined kind of form the growth discipline. Now the question about whether or not it just becomes part of good product teams or something like we've talked a lot about this with like Casey Winters from Pinterest grubhub and spent some time at Greylock as well. And he very much believes that. And I do feel like it's just part of being a good product professional in the sense of like we have these different tools in our tool belt and we need to understand when to use what tool for what type of problem. Right. And so sometimes the problem is very good for experimentation and for growth strategies. Sometimes we're going through much more of like an innovation exercise that doesn't necessarily land itself well to that. But at the same time, I think what we're seeing with all these disciplines is that they're going through a trend of hyper specialization if you look at design, engineering, all these kinds of things. And so even now there's like a new emerging segment of product management that's very focused on data focused tools like either internal or external. Right. And there's all of that specialization going on. So I think that's an argument for maybe this is just like one of those specializations that sort of sticks around. And so I don't know exactly how that plays out, but I do think if you're taking a bet as a product or marketing or any professional, it is to your advantage to understand the different types of problems that you might encounter in building and growing a software product and what are the sets of tools that fit with that problem because you're going to be a much more versatile and valuable professional and leader if you do understand sort of that range of the tool belt.

Yeah, I mean one of the questions that we get asked a lot from our portfolio companies and elsewhere is, you know, looking to make our first growth hire, what should the profile look like there? And I Think there's a fork in the road between marketing and product. But if you could assume that they're talking about product and building the product led growth function, how would you define what the ideal first hire looks like there? And I think less so background but more the underlying traits and skills that you would want someone like that to have.

Sean, you want to take a crack at this one first?

I think ultimately, I mean to some degree we have to first define what these we even mean, right? Because for a lot of people a growth person that they think of hiring may be somebody who comes in and sprinkles magic pixie dust on the product and magically grows it. That's just not a thing, right? So I think that if your goal from bringing in a growth person is somebody who can work out where the product is failing to grow, but that you could change and you could materially impact the growth rate of the product by tailoring either its go to market or the product experiences themselves, then you'd logically arrive at the conclusion that you need somebody who's very intellectually curious, right? So you need a very, very intellectually curious type of product person who effectively is going to go scouting across all the different surface areas of the experience. The go to market, the acquisition funnel, the activation funnel, the product experiences themselves and go looking for where the people are going missing, where the value is not being successfully captured by users. Then they're going to conduct a scientific search for ways to fix that and they're going to grow the product. If you think about what that type of person looks like, they look kind of product y, right? But they also look very intellectually curious. They are also very data oriented. They have a tendency to not be willing to stay inside a box. They don't tend to look at some specific thing like I own this feature or function of the software. They look at it like owning the overall experience. And they have a tendency to be change agents in terms of if they can see that the problem is somewhere that isn't directly inside their scope, then they will be out there inside the company pounding the pavement and beating the drum in order to change that. So they have a tendency to be people who operate outside of org structures in terms of willing to follow their nose when they work out what's going wrong.

I do think the starting point of the conversation is about a person and not a team. It's a setup for failure. Like one person just can't do that much ultimately at the end of the day. And you do need this hybrid of skill sets to truly move the needle. And so that's probably one of the top three most common things that we see people set up for failure. I think the second thing is that it's either set up under just purely the marketing function rather than the product function and they get siloed away from tech and product and design resources. And the third is that there's no like real CEO buy in around this is the time that we need to go do this and we need to establish this and we're going to give it the proper breathing room. You've got six months to establish, you know, your team. Right. I would actually argue for like a little bit longer, but yeah, the conversation has to be around a team and not a person, otherwise it's most likely going to tank.

Yeah, I mean, I would even take that a level higher and talk about establishing a growth model. It's one of the concepts I love that you guys teach as part of Reforge, but it's amazing. A lot of companies, they will start trying to move certain metrics without kind of having a holistic or complete understanding of how the product actually grows and what healthy versus unhealthy growth looks like. Tell me a little bit more about situations where you have seen growth models established really well within a company and maybe bring a couple of examples.

I think the interesting question in exercise for anybody listening to this, to go get four or five people in a room, give them all a blank piece of paper, ask them the question, how does our product grow? Have them draw it out or map it out about how they think it does privately, and then have everybody put it on the whiteboard and see what it looks like. Nine times out of ten, what you're going to end up with is very, very different pictures. What that means is that everybody is working from different mental models of how they think about how the business grows. And the problem with that is if that is the starting point, then any of the conversations that happen thereafter are going to be talking. It's basically people are talking a foreign language to each other and they don't even know it. Things like prioritization and what's going to influence growth the most. And all of these types of things is just incredibly difficult to have unless everybody's working from the same starting point about a hypothesis about how your product grows. The interesting thing is, in product, we tend to have these tools or these mechanisms for the other parts of product in terms of like, who is our target audience? Right. And what problem are we solving and what is our Value, prop and solution. And the weird part is, is when it comes to the topic of growth and how does our product grow? Typically the answer is like, well, we're just going to try a bunch of things and you know, whatever works is like, we'll keep going with it. And like that's not a hypothesis, right? Like that is not actually a model for how things grow. And so a model is really just about answering that question, how does your product grow? And you can answer it either in a qualitative or a quantitative way, but I think more than half the value actually comes from just answering it in a qual, qualitative way at least. The viewpoint that we take at Reforge is that the working framework to do that is to map it out in this concept of self reinforcing loops. And it's very simple. We have people, there's a bunch of different types of loops. If you look at all of the fastest growing companies that have driven those companies, that's not necessarily endpoint, but it's a good starting point to look at and learn and understand. Well, what are all of the different types of content loops? From a user generated content loop that drives something like Pinterest to a company generated content loop that drives a lead engine of HubSpot, to different types of viral loops, to all these other different types of loops and you can map out the few that you know is your hypothesis of how you think this product is going to grow. And more importantly, when you map out those loops, like, what is kind of the why behind it? So why is somebody generating content? Why is somebody inviting a friend? Why, like all the whys behind it? And that is the starting point for everything.

Yeah, I mean like growth loops are everywhere. Once you start to think about it that way, and you often find yourself in a conversation and you're talking through some specific experience and you go, oh my God, there's a loop there, there's another loop there. And it's pretty amazing when you get to that point. I would kind of break it down a little bit more simply though, to start with because people are like, I don't really know how I necessarily how to think about a loop and how to begin at a loop. I start with, okay, who is the person who should be loving your product? Right. How would you know if you have any of those people and how would you prove to yourself that those people who are the people who should love your product are getting the value that you think that they're getting? Right. Because I find that most of the Time, once you go in and you start looking at your data, you realize that, oh my God, that is happening at all. Right? You start to go, oh my God, Mike DeBose should absolutely love my product. And he appeared to fail on screen too. And he's now told all his friends that he hates my software. Right. So that you are the anti growth loop. Right. And so you start to kind of look at that and you realize, okay, well this is where people who are in my target market and should be the people loving my product, they are failing. So that is on you. You should fix that. But then once you've got that down pat, you go, okay, well where are more Mike DeBose. How do I get this person to tell other people about this? Like, how do I build a self reinforcing thing so that more of those target people bring in more of those other target people? And you can't have one without the other. They're yin and yang. Right. You cannot grow a product that is not a perfect fit for some identifiable group of people. But similarly, it doesn't just magically grow by itself. You need to build in loops that enable it to successfully grow.

I mean, so much of this is cultural. I think the concept of funnels are deeply embedded in companies, especially more mature ones. They've structured their entire. You have your acquisition team out in marketing, you might have retention, living in core product. An entirely different org, gold on different stuff, there's a conversion team sitting somewhere in the middle. I think the concept of loops is powerful in that it kind of throws that away and is a much more integrated approach to growth. But as far as team construction, if you're introducing that late in the game, how might orgs that are already structured around the concept of funnels think about kind of introducing that later in the game?

Yeah, I think in most orgs that are structured around funnels, what they'll find is if they take a small team and try to look at things through this lens, what they'll find is there are loops that exist within the product that just either like the final step or one step in the loop might not be like fully connected, or there's a piece that isn't like fully optimized and that is the result of having the teams kind of siloed and structured sort of in this funnel format in the sense that there isn't like anybody thinking about the connection of like whatever is coming out the other end of this thing. How is it being reinvested to like generate more of this? And so that tends to just be the starting point in the opportunity to find kind of low hanging fruit. And so if you do take the small team that is more of a mixture of product technology and a marketing focus and, and start to look at it through this lens, I think you tend to find these spots where it's like, oh, we kind of have this, but we just haven't connected this piece. And if we just insert this piece right, we all of a sudden have the start of a loop. I think the second thing that I naturally find is that the biggest things come from just first observing the natural behaviors of what your users are already doing and then productizing it in a loop format. So I'll give you three different examples from Reforge. So, for example, we run this application period. What we found in the early cohorts was that one person in the org was going to take the program and then they would try to get a bunch of colleagues to join, right? So, okay, that's easy. We productize that into a really easy viral loop in the application flow where we ask people to bring other colleagues with them. That worked. That amplified things greatly. Then what we saw in that same application flow was that we had a lot of people trying to find common connections with us on LinkedIn and have those people send recommendations to us or have their managers send recommendations to us. So then what we did is we productized that we asked them to insert their manager's email and vouch for them. And so then the manager comes in and vouches for them and then we ask them, are there any other team members that you have? And so that ends up creating a bunch of different things. We also see like people go through the program and then when they find something that's super relevant and applicable to them and their team, we find that they're trying to, you know, share snippets of it with other people in the org through like these very manual presentations or other types of things. We found a lot of this through customer feedback. And so we productize that into. These are all examples of viral loops, for example, but. And you can do this with any types of loops, but it was all about like observing sort of this natural behavior of what people were already doing and then finding ways to productize it. It not only delivered value to the user and amplified what they were doing and made it more frictionless, but it also delivered value to the business as well. And that's kind of like where the combination and where all the low hanging fruit comes in.

I think much of what you're mentioning rests upon an ability to actually learn both qualitatively and quantitatively about what's happening in a product right now. To your point, Sean, I think one of, I guess the mistakes or missteps I've seen in earlier stage companies is around analytics. I think there's a well understood concept of tech debt. I think analytics debt is an equally kind of common problem. And many assume that just going and buying amplitude or mixpanel or something solves all your problems in analytics. And it's much more than you look at. PM's workflow and instrumentation is just not part of it at all. Talk to me about analytics and I think reforge obviously should be best in class on ability to learn and productize from those learnings. But for companies that might be a step behind in analytics or really want to improve their ability to listen and learn, what advice might you have?

Yeah, well, I mean, I guess you're kind of hitting on a sore point for me. Like if I could do it all over again, for example, at Atlassian, then of the first hundred dollars I invested, I probably would have invested $80 in data because I realized that you literally cannot build a growth organization without like really clear, really good data and instrumentation because everything you build on top of it is just uncertain. And so it's utterly critical. I think as you mentioned though, one thing you can begin if you don't have very good instrumentation, you're not really in a great position at that point. You can begin with qualitative while you are busy doing quantitative. For example, if you integrate a full story and you look at 10 or 20 people doing whatever the product experience it is you're looking to try and improve, your heart will break. You'll see them all failing. Right? And it's not a perfectly quantitative view of what's going on, but it will give you a good place to start. But similarly, after that you are going to have to pick a tool, you're going to have to pick an amplitude or a segment or whatever else and you're going to have to go and instrument the key activities that are conducted in the software. And then you're going to have to set up some way in which the product teams and the different people that are enhancing the software are incentivized to make certain that the instrumentation is good. And I think product management of the future is really just where product teams do not celebrate shipping anything, right? If you do not allow a product team to celebrate shipping and you only allow them to celebrate if they can tell you what the impact was, then by definition they cannot celebrate until they've instrumented something correctly. You have to set up the system so that the system reinforces that the impact of something is the only important thing. And that will lead to what you're hoping to see in terms of instrumentation and data capability.

Yeah, I think Alex Schultz at Facebook took that one step further and said the growth team never rang the gong bell. It wasn't even impact. It was, you know, just how are we iterating on this and making it better? And I think that's very much the DNA of many of the successful growth teams we've seen.

I'll give you an example. I was talking to a customer literally yesterday. This person is a director level growth person at a company that we all know and everybody listening here would know. And this person literally in the interview said we were talking about a part of the reforged material that was more quantitative. And this person got there and was like, look, like I can't even get those numbers easily in a time fashion manner. So I don't even try. Okay, so like what's going on there? Right. We can unpack that in a number of different directions. But what a lot of companies don't realize is that talking about loops, well, there's this wheel of death loop going on where if you're not investing enough in the data infrastructure, then people either lose confidence in the data or there's too much friction to get to the data. And as a result, they don't use the data, which leads to lower investment in the data. And then it just spins to a point where you've got this thing sitting around that's like absolutely useless. And a lot of times people view these projects, it's a project mindset. We're just going to do this project, it's going to be three months long and it's done. And it's like, no, it's got to be a core part of your product process. And you have to put the right incentive mechanisms in place, as Sean was talking about, to make sure to reinforce it and support it over time. And then once you do that, then you end up with a lot of companies. I've seen you end up with the problem of the people with centralized data teams that view themselves as these gatekeepers of the data. And that's just as big of a problem because then they create themselves in their own bottleneck. What they really need to do is flip it and say, our purpose is to enable all of the teams with data and to enable the PMs with the tools that they can use and the data analysts with the tools that they can use and marketers and so on and so forth. You really got to look at this as a system eliminating friction. It's a growth problem in itself. But yeah, it's like something that we see over and over again. I don't think a lot of companies really know what's happening because your PM or your director of growth is not going to say the same thing that that person just told me, right? They're not going to stand up and say, well, I know I should use data here, but I didn't because it's just too hard, right? You're not going to hear. And so a lot of people just don't understand what's going on underneath the surface.

If I could do it again, I would do a bunch of different things. And so when I first started at Atlassian and we were starting the growth team, my focus, given that we had six months to live, was wins, right? And so I optim pace of learning and actually I don't even think it was even pace of learning, it was pace of likelihood of finding wins. And I don't regret that, but I do regret investing 80 of my dollars in that because effectively when you first begin with a growth team, you are just going to be shooting fish in barrel. It's always like that because all you have to do is go looking. And once you go looking, you realize, oh my God, did you know that X number of people drop off on step Z and you're like, oh my God, all I do is add a button and bam, there's a win. What that does is that actually gives you a return on investment that you can then choose what you next invest in. And do you pull another rabbit out of a hat or do investing in the platform that will enable you to sustainably pull rabbits out of hats for the rest of the life of the company. And I think we did some of that, just not enough of it. And it was not clear to me early enough how much data is literally unstoppable. Like if you don't have it, you simply cannot win. And the thing is you need the data because you need enough data over time, right? So an experiment system, any day you can pick one, stand one up and run a project and before you know it, it'll be running any AB measurement system. You can again do that too, right? But long longitudinal behavioral data, which is critical given growth loop or growth model, if you don't have it, you will never have it. So you're just critical to get that in place.

This is another one I pulled from Alex Schultz, who I was fortunate to spend time with earlier in my career in growth. And he said there was an arc of how teams treated data and there were four stages. I think first was just not having any data at all, just flying blind. Two was starting to have data but not trusting it. Three was having data, trusting it, but using it totally improperly. And four was things working harmoniously. And it was rare for companies to make it to stage four.

Yeah, I think one more data point. This is Ludo Antonoff, who was the head of growth engineering at Pinterest but now is at Lyft. They had a very specific discipline on the Pinterest growth theme, which I would argue is probably in one of the top growth teams out there right now, where 60% of their time and energy and resources goes into the infrastructure, the data and the experiment infrastructure. 40% actually goes to the actual experiments. And he and I were talking a lot about how when he puts that slide in front of a lot of other executives, a lot of companies, how like eye opening that is for people and how surprised people are at that shift of investment. But once again, it's this thing that enables you to spend the 40% on the good things and not flying blind.

Yeah, it's investing in the infrastructure layer. And I think something that it's actually a good opportunity to shift gears to the hot topic of paid marketing. Because I think to your point, Sean, on just needing to find quick wins at the start, I think for many businesses that are well resourced with VC funding or otherwise, pretty easy to just go turn on an acquisition channel, say Facebook, which is fair automated these days, and just start driving volume from there. And we talk about health and durability of growth. I think too often companies go too far down this path and it actually can mask underlying health of growth. At Reforge, it's interesting you actually don't spend too much time on paid marketing. I'm curious for your point of view on how companies should think about using it responsibly as a tool versus being overly dependent and treating it as a drug.

Yeah, I mean the main reason we haven't really focused on it is just limited time, resources and only so many things. And we do think viewing things from a product centered out viewpoint is probably the best starting point for anybody on this. But we do very much view anything around PA as being hopefully an amplifier to one or other set of core loops within your model versus it being the core of the model. And we've seen this time and time again when it's the core of the model, it just ends, ends up saturating and sort of hitting the ceiling very quickly. I think the bigger thing is that most companies aren't really good at anticipating where that ceiling is and as a result planting the seeds far and ahead of time to figure out whatever that next S curve is or whatever is going to raise that ceiling. And because establishing a new growth Hooper, one of these initiatives is not like, you know, I can draw it on a piece of paper, but it probably takes, it takes a long time to establish and optimize and make one of these things, things super meaningful and super core to the business. And so I think that's one of the bigger things. We just talk a lot about it being an amplifier to the core and more importantly also just thinking and anticipating kind of where the ceiling is at through a number of techniques of burst testing and a bunch of other things that you can do to at least predict a little bit of where it's going.

Yeah, you have to be really careful. Channel saturation is a thing. Every given channel that you're in which isn't yours is only ever going to go in one direction and that is get more expensive. Right. And usually exponentially so. And surprisingly, usually at the end of quarter when you're trying to make revenue or whatever, suddenly everyone else is competing in the same channels. Right. It's just the only channel that you ultimately control is the experience of your existing users and whether or not you can successfully scale them to acquire more. It's obviously true that for Atlassian this was just like incredible killer secret sauce that I last saw. The data. About 80% of all new logos were acquired from no known source organically. Imagine being you're at a billion dollar run rate and you're still acquiring 80% of all new logos with no contact with them. Right. And they're not by any paid channel. They're not anything. It's an unstoppable ball of revenue rolling down a hill when you have successfully engaged your own user base in that way. Right. To spread your product from person to person. And the thing that Brian is pointing out is so important. It's like it takes a lot of concentration and organizational sacrifice to successfully do that.

That.

Right. And so you need to spend years and time and thought and everything else to build this sustainable thing. It's worth it, but it does not happen instantaneously. And if you delay Starting it, you won't get the fruits of that. It's a bit like, you know, you've got to plant the seed and let it grow in order to successfully get there. And when instead you're at end of quarter using paid channels or finding new paid channels or whatever, then the organization doesn't get the time and space to focus on truly doing that.

Yeah, I mean, there are also certain business models that won't necessarily support that type of work that maybe are dependent on third parties, and those are maybe ones that it's more of a grim reality as far as what growth at scale looks like. I think one of the things you're saying resonates from my early time at Stitch Fix. I mean, Stitch Fix actually is in kind of a world where it's typically highly dependent on paid marketing. But in the early days, it was very little of that. It was just focusing on the right customer that was underserved in a part of the country. That word of mouth just spread really far and wide about this. And really the core capability that the company built was around retention, demand for forecasting, algorithms and all that. And so by the time we started scaling on paid later, there was already a DNA in the company that was likely to make us pretty effective at it. I think even with that, it is rare that a company, I think, sustains differentiation based on paid marketing in the long run. And so I agree. I think types of business models where it could be an accelerant versus a crutch. I actually think it's a competency that companies should think about building out and getting great at. But. But I mean, the point you named Brian, which is really around just understanding marginal CAC and where you are in the saturation curve, that's something that it's actually kind of hard to understand.

It is hard. Yeah, yeah. I say it, we talk about it, and it sounds easy, but it is hard. But it's one of the important hard things to do and try to get better at over time. For sure.

Cool. When you put it that way. Effectively, you're kind of defining phases of an organization or phases of a product. Right. And something that certainly resonates with me in terms of phase one of any given product is building a product that people are going to love and really love. Right. And so you're talking about Stitch Fix and my wife's a customer and like absolutely loves it. Right. And has been a customer for a long period of time. Right. And so effectively, in phase one, that company successfully engaged exactly that word of mouth. Engine and it was that that allowed phase two to exist. Phase two, when you used basically you poured a paid acquisition on top like it's just fuel on top of the fire that's already burning. Right. And then eventually you're big enough that you don't need that anymore. You're now a self sustaining system where you can start to reduce your dependency on paid. It's when you put it in too early, when there is not a sustainable growth lo not a sustainable lovable product that it deceives you into believing you have found a really great product and you have not yet.

The reverse sequencing actually makes me particularly nervous in companies that have crack paid and then they say oh, phase two is going and figuring out organic.

Yeah, isn't that awesome?

SEO is not something you just turn on in a matter of weeks I'm.

Just going to go and figure out how to make my customers love this product. And I'm like what? You're already at this scale and now's the time to work out how to make people love my product.

Yeah. So I guess just to close out as I kicked off with there's many folks out there that are thinking about either hiring or investing the next chapter of their career in doing growth, either down the marketing path or product path. What career advice would you have for folks who are just getting started in it or are at a trajectory in their growth career? To both you guys?

Yeah, I mean it really depends on the person. I wish I had three piece for I I have people patience and I guess we could call it positioning. We'll label that people. That's self explanatory. I think everybody knows to optimize for people. Patience is one of these things that I look back at my past 10 or 15 years and I look back at the most rewarding experiences and then the ones where I've been most patient and really dug in and really gone beneath the surface and then the ones that I look back and I kind of regret a little bit it I'm like I wish I was just a little bit more patient because if I had been I think like we really would have gotten to the next level. Whether that's my time founding some startups or some things that I was working on at HubSpot, I think that patience is actually just like really key because like a lot of times especially in larger orgs and you like see something that you really want to do that you really believe in a lot of time it just comes down to timing and waiting for that right time. Time that it fits into, like, the broader plan of whether it's the org or the team. And when that time comes around, that's when you can like, really nail it. But. But I see a lot of people being like, really impatient and end up just zigzagging all over the place. And it's really hard to build what I would call like a meaningful body of work if you're constantly zigzagging, jumping like from one thing to the next. I think the third is, like, positioning is, is really thinking about how you're not only just positioning yourself inside your current company, but, but the position of where your company is at in its growth trajectory. And so I think the best point is when you enter a company that is growing faster than they can actually hire. And the reason for that is, is that as a result, you will get opportunities to step into things that you might not otherwise get elsewhere, which is really interesting. And then actually which projects you take on. I tend to find a lot of times the most impactful ones are the ones that, that aren't the sexiest. They're the ones that nobody really wants to touch. And a lot of times the person who steps up and it just takes on this, like, really hairy thing that is really impactful, but just not sexy at all, is you position yourself as being in the flow of value, like within the company. And just once again, you end up creating leverage for yourself by like exposing yourself to a bunch of opportunities that you might not otherwise. And so, so those kind of three things is like, people, patience and positioning is how I would think about my past 10 or 15 years and what I would tell myself.

Yeah, riffing a little bit on Brian's view. I think a lot of this comes down to locus of control, right? You have to decide, does your career happen to you or do you happen to your career? And what I mean by that is you have to be willing to run towards burning buildings. When you're inside a business and you observe that something is broken, right? And it is key to the ability of the business to succeed or not, you have to be fearless and go, well, I know nothing about that thing, but I'm going to go and learn, right? And similarly, right, if you can't figure something out and you're not certain why this thing is broken, you have to be fearless and you have to be willing to do the hard yards and think about it and get all the qualitative data and everything else you need to do in order to successfully grow as a person. My career has, I Guess Brian said zigged and zagged. But I've done a lot of different things and each one of them has become an arrow in my quiver. Right. And it means that when I look at a business, I don't have to worry about whether or not the problem is in marketing. I'm happy to go and talk marketing. I can talk marketing to marketing people, I can talk product to producting people, I can talk engineering to engineering people and sales to sales people. Right. All of those things are just capabilities that are needed in the end to solve the problem that matters to me, which is why do some product companies grow and others not? What does it take to set fire to a product company and have it truly deliver value to the world? What would it take to build a truly great product company which can measure its impact on the world in terms of millions of people touched and millions of hours saved and other values like that? So when you let yourself follow your nose towards the things that you desire and don't take no for an answer, you just grab onto any opportunity as it comes up and you successfully crush it. That's when doors tend to open.

Yeah. There's two types of zigging and zagging. There's the zigging and zagging where you're jumping from one thing to the next because you start a project, it looks nice at the beginning and then you hit a wall. Whether it's like, I'm having trouble getting buy in or I can't get the data. And so you're like, then you just jump to the next thing. But then there's the type of zig zagging which you're talking about, which is like you do this thing, you dig in, you solve it, you learn a ton and then you jump to the next thing and you repeat the same process and just trust in yourself that I don't know this thing. But I'm going to figure it out. Right. And I'm going to get better at figuring things out the more I do this. But you have to give yourself that time and that patience to actually figure it out and work through the hard stuff.

Yeah. This is a common thread that I see across growth people is like this insatiable curiosity and just desire to learn as quickly as possible combined with also so a willingness to fail. I think that's what I see across some of the most successful growth people. I think it actually can be a bit tricky or dangerous in that that will often lead you out of a specific role or company faster than you might otherwise intend. And Actually a lot of our peers have taken on roles in growth consulting because it actually, I mean, what a great way to go and just totally accelerate your learning and get a broader surface area of learning as well. I do think that curiosity combined with an ability to reflect fail and also operate and learn from that failure, I think is key.

Yeah.

To be truly impactful on companies, you have to be able to do multiple different things. So it's not just simply create a thing, it is get buy in of the company to go do the thing. It is get momentum of the company towards driving that thing into said market. Right. There are many different things that it takes to truly be successful. So the toolkit, the intellectual firepower and the toolkit to go do it. And even having done it are not by themselves enough. Right. As you develop in your career, you still need to gain influence capabilities. You need to gain outside information that you can bring in to leverage to get better influence. There are all these different things you collect as a human and just a random anecdote. Years ago, I ended up one of the primary authors of the Atlassian privacy policy. And this was an unbelievably painful process, writing Atlassian's privacy policy. And what on earth is a growth guy doing writing a privacy policy for a major public company? That makes no sense. Right. But actually two things happened out of that. One was, was that I realized that the flip side of what I cared about, which was this data that enabled me to do my job, was that this data meant that I was a steward of this important context for people. I was a steward of the confidentiality of the things that they were doing with our software. The yin to my yang was being willing to do that as well. Right. I wasn't trying to drive, just my agenda. My agenda was I did want to grow this company. I did need the data in order to do that. But a side effect of that meant that there was potential privacy implications and everything else. So we had to be willing to, to walk the walk and talk the talk. If I was going to talk the talk, I had to walk the walk. And so the privacy policy was it. And that was a really interesting experience. I doubt I'll ever have to do that again, but it was fun.

Well, guys, this has been great. I mean, one of the things I really love in this and also talking with you guys just in general is we really don't. Despite the popularization of the growth hacking term, which I think has become less popular nowadays, we don't sit around and talk about tips and tricks on this stuff. It's much more about the fundamentals. And I think Reforge is really, really you practice what you preach on that one. So thanks for joining. Really could have kept going for a while. So for those in our community who are listening, where could they find you after this?

Yeah, Reforged.com, if you're interested in joining the community and being part of our programs, you can sign up for a notification list there. I also blog not nearly as often as I would like@brianbelfour.com and I'm off social media. That was my thing for this year. But those are the two best places.

Pretty cool. I wish I could do that. So I blog even less regularly than Brian does@seanclaus.com, com but that's probably the best.

Cool. Thanks guys.

Thanks.