How I Built This: 100M+ AI Startup

1. 0:00 – 1:26

   The 15-step framework to build a $100M+ startup (and why it’s hard)

   1. MAMatt Arbesfeld0:00

      If you wanna start a company, how do you do that? And so we're gonna be sharing 15 steps to go from nothing to building a company worth hundreds of millions of dollars.

   2. AGAakash Gupta0:08

      What has been the lowest point in your journey?

   3. MAMatt Arbesfeld0:13

      There's just a constant stress of starting a company, 'cause you're just bombarded with problems every single day. The hardest times are just the ongoing grind of dealing with problems.

   4. AGAakash Gupta0:22

      Now I think we're into step 10, raise more money and uplevel your team. Talk to us about that.

   5. MAMatt Arbesfeld0:27

      You might decide to just bootstrap, and n- nowadays I'm hearing you have, like, a one-person company that's 25 million of revenue because of all these AI tools. So you might not need to raise any money.

   6. AGAakash Gupta0:37

      If you just wanna build a 1 to $2 million business, it's not venture scale to begin with. Like, venture scale, what are we talking about when we say what is venture scale?

   7. MAMatt Arbesfeld0:44

      So I even wonder, if you're in a competitive market, it can be even hard to build a $2 million business because-

   8. AGAakash Gupta0:52

      Really quickly, I think a crazy stat is that more than 50% of you listening are not subscribed. If you can subscribe on YouTube, follow on Apple or Spotify podcasts, my commitment to you is that we'll continue to make this content better and better. And now on to today's episode. If you wanna build an AI startup that's worth hundreds of millions of dollars, there's really no better way than learning from someone who has. Today we have Matt Arbesfeld. He created LogRocket. Welcome.

   9. MAMatt Arbesfeld1:23

      Awesome. Great to be here, Aakash. Thanks for having me on.

2. 1:26 – 2:25

   Step 1: Build projects for years—your ‘founder origin story’ matters

   1. AGAakash Gupta1:26

      So you and I have come up with 15 steps to creating this successful AI startup. What is step one?

   2. MAMatt Arbesfeld1:34

      So step one, and a lot of people ask me, "Oh, how do you start a company? What should I do?" And I've never seen someone start a company if they haven't spent a lot of their life just working on projects on their own and trying stuff. And so my origins of starting a company probably go back to elementary school, when my now co-founder, we were building projects for classmates. We were trading food at lunch. We were, uh, kind of, uh, uh, creating trouble, uh, to try to build businesses back then, all the way until college, when, uh, another friend and I were building mobile games and, and selling those online. So I think a lot comes down to just trying stuff and building things, and, and that really helps you develop your skills as a creator and as a entrepreneur.

3. 2:25 – 3:20

   Failures and wins: learning from a dead app and a viral mobile game

   1. AGAakash Gupta2:25

      Are there any specific things that you guys spectacularly failed or spectacularly succeeded at? [laughs]

   2. MAMatt Arbesfeld2:32

      So I, I'd say, uh, probably my most failed app, and we spent, like, a year working on this thing. This is in college, like, nights and weekends, no social life. The app was called Cluck, and the idea was I was this, like, socially awkward college kid. I didn't wanna call people, so I wanted to enter a real-time texting conversation with them instead. So instead of calling, you know, I could cluck you, Aakash, and then we would be texting in real time back and forth, and you couldn't leave the conversation. But a completely useless idea. None of my friends would even use it. We launched and we had, like, five total users, so that was, that was a complete failure. But then we, we had a mobile game in college that we reached a few million users, uh, that. So you have some wins, you have some losses. It's, it's hard to have 100% batting average.

4. 3:20 – 4:31

   Distribution lessons: how the game got traction (and why virality starts intentionally)

   1. AGAakash Gupta3:20

      What did you learn from building a mobile game to a million plus users? That's awesome.

   2. MAMatt Arbesfeld3:25

      A, a lot. Uh, so definitely a lot of work went into building. The, you know, we were spending all summer after work working on this thing. Probably the reason we got that much traction was we spent a lot of time on distribution. So as we were launching, this was back in the day of press being a big thing, so I, I worked with all these, uh, game writers and publishers and Twitter influencers to, to share the game ahead of time, so they got excited about it. And they would publish on their, on their, uh, on their articles in the journals, and that ended up getting some initial traction. And then I think the game was just fun, so people would share it with their friends, and they liked playing it, so it went a bit viral, uh, after that kind of big launch.

   3. AGAakash Gupta4:14

      That's awesome, and I think that's a really important part of the journey for people is you're not just gonna luck into building something successful. Obviously, Matt has taken a lot of sw- swings of the bat, and he even had a pretty successful product, so that's awesome. What is step two?

5. 4:31 – 6:49

   Step 2: Develop hard skills (and learn by building what you care about)

   1. MAMatt Arbesfeld4:31

      So step two, and this is probably intermixed with step one, so maybe it's 1A and 1B, would be you need some skills, right? Like, what are you one of the best at in the world that you can bring to the world in your company? And so I spent a lot of time since middle school and, and high school writing code, but also I've, I've loved design and, and visual things, so I've spent a lot of time, um, designing 3D models and creating graphics for school projects. So, um, I think kind of those were two of my passions growing up and, and have kind of come to fruition with the company I'm building now. But, um, kind of finding something you care a lot about, going deep, being an expert in that field, and that forms the basis usually of whatever company you start.

   2. AGAakash Gupta5:22

      Yeah. I think that there's a lot of people who wanna just vibe code their way [laughs] to a successful company, but without some hard skills, it sounds like for you across coding and design, right, basically the two most important parts of building software, you had developed some hard skills along the way.

   3. MAMatt Arbesfeld5:41

      Yeah. Um, um, who knows? Maybe now the vibe coding will be a, a top skill. So, uh, you, if you go deep in that, you could be an expert there. But, but yeah, I think kind of having some fundamental skill that you're really good at and can, can, can excel in goes a long way.

   4. AGAakash Gupta5:58

      If you were to advise somebody, let's say they're basically starting at zero in terms of coding and design, but they're smart and hardworking. Where do they start?

   5. MAMatt Arbesfeld6:06

      I'd say start with one A. Find a project you care about that will help you or your family or your friends, and just work on it. And the first time you work on something, it's gonna be really painful and challenging, but you're... Fail, you have to learn how to build the thing. It won't work. It'll be really frustrating. But that, I found, is the best way to, to learn stuff, is just apply it to something you care about and, and launch and, and, and fail.

   6. AGAakash Gupta6:32

      100%. I imagine, like, just the trials and tribulations of scaling a game to so many users, you're learning so quickly just based on the feedback you're getting from users, and you need to build this, so you're learning how to code this, or you need to improve the design here, and so you're learning that element of design.

6. 6:49 – 9:08

   The MIT “maker culture” advantage: environment that normalizes building and failing

   1. MAMatt Arbesfeld6:49

      And I, I even remember back at MIT, you're forced to take these humanities class, 'cause they wanna culture you in things beyond just computer science. So I took a poetry class, and for the final project, most people would, like, analyze a poem, but instead I built an app that automatically could help you analyze the poem and, and, uh, annotate what's going on. I think my teacher was super frustrated that we were, like, applying computer science to a humanities problem, but just finding opportunities to build and do things I enjoyed in all walks of life, I think kind of, uh, helped me develop the skills for building and launching things.

   2. AGAakash Gupta7:26

      That's a great point. You started at MIT. Spencer, the last CEO we just had on, talking about a similar story, started at MIT. There's something there that they're really encouraging you guys to work hard, do side projects, and build. What would you attribute, what did you learn from that MIT culture?

   3. MAMatt Arbesfeld7:43

      Uh, for sure. I think there are a lot... Yeah. Spencer's built a great company. There are a few other. Braze was MIT Segments. So, um, definitely there's a culture of building and starting things. I think one is they definitely select for that. Actually, when you apply, you can submit a maker portfolio now, which is basically things you've built in high school and middle school. And so clearly they encourage that. And then a lot of the dorms and culture at MIT is around building things. So there's this one dorm there called East Campus where at the beginning of every school year, they build basically an amuse- an amusement park. So they build these huge contraptions, like a big swing, swinging arc thing, and these giant slides. I'm sure it's super dangerous, but-

   4. AGAakash Gupta8:26

      [laughs]

   5. MAMatt Arbesfeld8:26

      ... uh, they definitely encourage and support the students creating stuff and, and failing and building things. Uh, that's just kind of inherent in the culture there at MIT.

   6. AGAakash Gupta8:36

      That's super cool. I think that's definitely an important part of it that maybe not everybody can work backwards from, but I think if you're in high school or something, definitely having aspirations to go to MIT and reverse engineer what it is that's required to get in there, like you said, a maker portfolio, probably a really, really high SAT score, and a good GPA, you should try to do that, actually. I think it's actually worth it, because when I'm talking to different CEOs [laughs] like Matt, there's a crazy percentage of them coming from these small amount of schools. So what's the next step?

7. 9:08 – 12:50

   Step 3: Find a problem domain through real exposure—Meteor to LogRocket insight

   1. MAMatt Arbesfeld9:08

      So yeah, I'd say next step is once you know how to build stuff, and maybe you failed a bunch of times or had some small successes, you probably have been exposed to a lot of different problem domains that, where you've noticed problems and you saw opportunities. So, um, in my case, I spent a lot of time at a company called Meteor that was building this front-end JavaScript framework. So I spent a lot of time in the world of front end and React and Angular, learning about this whole world and seeing all the problems that were present in front-end development and building apps. And that experience led me to see the opportunity for LogRocket and, and see that instead of just a mobile game or some poetry app, there's a huge opportunity for every developer who's building these React apps to build a better solution. And so yeah, I think it's important to, as you're going through work and life, kind of getting exposed to these problem areas and digging deep to understand them and, and find opportunities to build companies.

   2. AGAakash Gupta10:09

      Today's episode is brought to you by Miro. Let me ask you something. How many tools are you juggling just to get a single project across the finish line? One for brainstorming, another for planning, something else for tracking tickets. That's where Miro comes in. It becomes an all-in-one collaboration workspace. Whether you're consolidating user research from several interviews, developing and synthesizing product briefs or a wire frame, or project managing development, Miro brings everyone into the same space. It's fast, intuitive, and fully loaded with features like project templates, two-way Jira sync, and integration with software like draw.io and PlantUML. Miro's AI features can be used to synthesize elements in a board to develop a ready-to-review product requirements document in seconds. If you're tired of tab overload and scattered workflows, try Miro. Head to miro.com and see why over 90 million users choose Miro to guide from idea to outcome. Today's episode is brought to you by Amplitude. Building great digital products is hard. You know that better than anyone. Getting teams aligned, measuring what matters, and scaling your product strategy isn't easy. But what if you had a clear framework to guide your next steps? That's exactly what Amplitude built. They studied the best product teams to understand what really drives impact, and turned those insights into the digital experience maturity assessment. In two minutes, you'll be able to see where your team stands and what you can improve to build better products faster. Click the link in the caption to take the free assessment and get a clear path to product growth. Yeah, this is probably the one that all the product managers are nodding their head on, because understanding the problem space, having a deep, intimate understanding, I think is usually the core of how you go about building great solutions. So tell us a little bit more about the problem space you narrowed in on for LogRocket, and how you kind of identified and circled, "Okay, this is the space we're focused on."

   3. MAMatt Arbesfeld12:04

      So yeah, if you think back to-... apps 10 years ago, they were generally super simple. You'd have a, a static webpage, and you'd click a button, and then the next page would load. Nowadays, tons of stuff is happening. There's animations, and th- you know, drop-downs, and very interactive UIs, and that was just starting 10 years ago with tools like React and all these new frameworks, Next.js. So I saw there were all these companies that were building back-end and infrastructure monitoring tools, but no one thinking about how you understand the user experience, and the front end, and the product experience. And so, saw that there's sort of a big gap in that understanding, and that's what led us to start LogRocket, uh, back 10 years ago.

8. 12:50 – 16:34

   Step 4–5: MVP scope and the first big launch (Hacker News momentum)

   1. AGAakash Gupta12:50

      So that brings us to step four, build an MVP. Talk to us about how you scoped out that MVP.

   2. MAMatt Arbesfeld12:56

      You know, it wasn't such a deliberate process, I'd say. My co-founder and I, we were living together, and we would just come home and be like, "Hey, let's work on something," and then we'd kinda throw ideas at the wall and, and try stuff. And we came up with this idea to, where you could capture videos of users using your site. So, uh, we didn't know this was possible, but we sort of were hacking at this, and we realized you can actually track everything that users are doing and see the visual. And so when there's a problem or there's an issue, we could actually go play back and see what happens to that customer. So we started working on that, and, and we launched, um, uh, later that summer. But sort of that initial idea was more kind of seeing that this technology was now possible and matching to the problem that we had identified.

   3. AGAakash Gupta13:44

      Okay. And how did you guys launch, which I believe is our next step?

   4. MAMatt Arbesfeld13:51

      Yeah. So we, we were spending all, yeah, all summer working on this thing, kind of, uh, no social life, just coming back after work, uh, a- and, and working hard. And then at the end of the summer, and this was when Hacker News was... Like, the Reddit was just kinda starting, but Hacker News was the big place people... You'd launch and hear about products. And so, yeah, I remember we would, we were planning this launch for weeks, and we told all our friends like, "Hey, it's launching this time. Be ready to up vote it on Hacker News." And so we posted, and we had all our friends up vote, and tons of people commenting and, and responding to what we'd built. And I just remember that day, like, we got probably thousands of signups. I was like, "Wow, we have something real here." Like, it was just a beta list. We didn't even have a public product that you could la- that you could sign up for, but we had, yeah, thousands of people, big companies too, of people asking for, for access. And, um, that was a super exciting moment where you're like, "Wow, we... There's something, there's something real here that people want."

   5. AGAakash Gupta14:55

      Wow, because most people, that's not how the [laughs] launch goes. I've worked with plenty of people, and it's just a couple, you know, couple chirps or something like that on Product Hunt that they get. How did you guys kind of... I guess you learned a lot probably from your gaming days, but what were the specifics to kinda make your launch do so well on these platforms and get so much inbound interest?

   6. MAMatt Arbesfeld15:18

      Yeah, and I think it is definitely harder nowadays. There's so much more noise everywhere. But I'd say, like, a similarity between the game and our, our LogRocket launch is you need this kind of mass of initial people liking and sharing something to get kind of y- off the ground, but then the product has to be great. 'Cause if you... You could have all the, you know, all your friends talk about it, but if no one actually cares in the real world, it's just gonna die, like my app Cluck. But if you build something that is, like, new and unique, and you have a interesting take on the world, that's when you can go from that initial, like, set of, yeah, like game writers or your friends and, and family, and then people start... It kinda goes viral. People like it and share it, and they comment, and they tell you how, how bad your idea is. But th- all that stuff helps propel it to, um, to get the, kinda that, that initial hit.

   7. AGAakash Gupta16:14

      And at the beginning, was it just still you and your co-founder, the two of you?

   8. MAMatt Arbesfeld16:19

      Yeah, and we were living with someone else, so he was kind of a contractor for us building some front end, but it was just, just the two of us, um, for a while. And then we brought on, um, our first, yeah, first engineering leader probably about a year in to the company.

9. 16:34 – 21:15

   Step 6: Raising the first money—traction, student funds, and warm intros

   1. AGAakash Gupta16:34

      Okay. So step six is raise funding. When did that happen, and how did that go?

   2. MAMatt Arbesfeld16:40

      Yeah. So we, we launched our MVP. Actually, we were between junior and senior year of college, so we were... Uh, I remember at the end, end of the summer, I was like, "Oh, we should drop out of school." Like, "We have something here." And my co-founder's parents were like, "No, you have to go back to school." So we ended up going back to school as we were... So I, I remember I was, like, in the back of class coding, not paying atten- attention at all to what the teacher was doing. Um, so over that kind of semester, our revenue just keep, kept growing. I think we got to maybe 5,000 a month of revenue. And one of my friends was like, "Oh, wow, you know you can actually just, like, go raise money for this? Like, you can probably raise a million, $2 million." And I was like, "Really?" Like, I had no idea that you could just, uh, go out and raise money and that was just a thing you could do. So, um, we ended up putting together a deck. My friend introduced me to a few people, and one of the first meetings we had was with, uh, Matrix Partners. And we shared our revenue, and what we were building, and the vision with this deck, and they ended up investing, maybe it was, like, a half million in the company at that point, and that, that sort of set us up to be able to work on this for the next couple years. Um, so, um, yeah, it was really just about going out there, and having that kind of initial traction and revenue I think really helped us sh- show that we could build something great.

   3. AGAakash Gupta18:07

      Yeah. If two students from MIT are already at-It sounds like 60,000 ARR, you wanna inject money in to see, like, where they can take it. So that's awesome. And you said your friend and they introduced you to Matrix. If somebody wanted to kind of reverse engineer this process, like how did you meet this friend? Like is it a s- is it an outcome of being in that MIT environment? How can other people kinda engineer that for themselves?

   4. MAMatt Arbesfeld18:32

      Yeah. So, and I think this is probably accessible to most students. So our first investors actually before that were, uh... They were called Rough Draft Ventures and Dorm Room Funds. I think they're pretty funny. Uh, Dorm Room Funds especially is a fu- funny name. But they basically just invest in student-run startups, maybe like 25K. So I- I've seen a lot of these kind of programs that, especially if you're a student, you can go and, and work with these entrepreneurship clubs and, and raise money from those groups, and they usually are connected to all these different groups. You also have Y Combinator, um, if you're later, or various accelerators that maybe are even before you have a product, they invest just in you as a team. So I think that's a good place to get started. Um, if you really have no connections at all, you could email me, email other entrepreneurs. Like, we wanna give back and help people who are just starting. So if you have something that's working like that, we're more than happy to introduce you to, to investors. So if you have nowhere to start, you just email people. If you have traction, say you have revenue, and, and usually they'll listen. That's their job, to invest.

   5. AGAakash Gupta19:42

      That makes sense. I think the number one key here, takeaway for folks is get some traction with revenue, right? Find a problem that's actually working. And I think for you, probably also getting revenue in the B2B space is a little bit harder than B2C. So if you, once you're hitting something like 5,000 MRR B2B, I think people's ears and eyes are definitely gonna perk up. And when did you start paying yourself along this journey?

   6. MAMatt Arbesfeld20:07

      Yeah. That, so that was, yeah, after we had raised that initial round, I think we paid ourselves maybe 50, 60K a year. You know, just to, enough to live off of, but not, obviously you're not making tons of money off of that salary, but just enough that you're not worried about paying rent and you have enough, like ramen, to, to, to buy ramen. Um, and actually my co-founder and I lived together, so that was great. We, we could share in the expenses there and, and the office. There was no... We would go from the office back to the house, and it was continuing to be the office. So it's kinda nice to be just immersed in the company as you're starting the business.

   7. AGAakash Gupta20:42

      Definitely hardcore MIT style. And did you finish college?

   8. MAMatt Arbesfeld20:48

      Uh, technically, yes. So technically I had enough credits after that, like three and a half semesters to, to finish college. And then my co-founder, I think he had to take like five gym classes in the last semester to, to get enough. So, so we both did finish college, so we're, uh, we, we have degrees, uh, b- uh, bachelor's degrees. I did drop out of my master's though. So-

   9. AGAakash Gupta21:08

      Okay

   10. MAMatt Arbesfeld21:08

       ... uh, I'm a bit, I can say I'm both a dropout and a graduate.

   11. AGAakash Gupta21:12

       [laughs] All right. So what's step seven?

10. 21:15 – 23:44

    Step 7: Build a durable GTM channel—LogRocket’s SEO/content engine

    1. MAMatt Arbesfeld21:15

       Step seven. So we were at the, yeah, five, you know, maybe grew to like 10,000 MRR, and you realize you can't just keep launching, right? Like you can only launch once. Maybe you can launch twice or, or three times if people don't see it the first time. But you really need a durable way to acquire new customers for them to find you. And so, uh, as I was coding and building the product, my co-founder, who's also technical, was building our blog. So we identified that our customers, frontend developers, had really no place to ask questions about how to do stuff using the frontend technologies that was a, a well-known trusted source of information. And so he started writing blog posts, and we realized they'd quickly rank on Google and people would find them, and we were getting tons of traffic through these blog posts. So that's where, how we grew the initial go-to-market channel was scaling this blog and content engine, uh, for frontend developers.

    2. AGAakash Gupta22:15

       And how did you guys approach, um, this? Like was he really like religious about looking in analytics, just religious about quality of content, lots of volume? What were like the keys to your SEO strategy?

    3. MAMatt Arbesfeld22:31

       Yeah. I think the fundamental thing to any content is quality. No o- any- no o- no one wants to go to a blog post and say like, "Oh, the thing doesn't answer my question," or it's poorly written or it's wrong. Like that's the death knell of content. So quality is super important and... But, you know, on top of that, you have to understand where people are finding you. So for us, it was through Google. People... This was back before ChatGPT, and you would just, you know, ask any question in ChatGPT. People would search up like, "How do I build so and so with React?" Or "How do I... What, what charting libraries should I use to build my app?" And so we knew all these search terms that people were looking for, and we'd write posts that address those, those topics. And I think at our, at some point we were writing maybe 200 posts a month, like very high volume, but keeping the quality, uh, high.

    4. AGAakash Gupta23:23

       Wow. That's an enormous amount of posts. So to cr- create quality posts, it takes several hours. So if you, it sounds like he was on the mega grind, like just creating a huge amount of content, and that's how you really succeed in step seven of building a GTM channel. So step eight, grow the team. Talk to us about how you grew the team.

11. 23:44 – 31:23

    Step 8–9: Scaling the org—recruiting systems, culture, and hiring executives

    1. MAMatt Arbesfeld23:44

       Yeah. So the, I'd say I, I've seen a plenty of companies die that the founders are, are great and they can build stuff, but to, to build something great, you can't just do it as 2, 3, 4 people. You need a, you need a great team, an amazing team. And so, um, recruiting becomes one of your primary jobs as the founder or as a leader in the company. And so we were very lucky to recruit some incredible people throughout the whole lifetime of the company. And so we spent a lot of timeThink about recruiting and how to find people and how to reach them and how to close them and, um, building just a great culture that when people would visit the office, they would... You, you wanna be in that environment and, and spend time with everyone there. So we spent a lot of time, like, building recruiting bots and, um, follow-up machines that would kind of keep people in, in our loop and, and also just, you know, hand-to-hand, like, meeting people, getting coffee, um, uh, finding great people to, to work with. And it's a self-fulfilling prophecy. Like, once you have a team of 10 great people, the 11th person who, who sees that great team, they, they wanna join, like, they're compelled to join. And so just helped really build a, an amazing initial culture for us.

    2. AGAakash Gupta25:01

       And you guys stayed in Boston, right?

    3. MAMatt Arbesfeld25:04

       We stayed in Boston, which I think was great for us because Boston's has really good technology people, but there's... If you think of what are the great companies that were built here, there's not a ton of amazing, interesting companies like you might have in the Valley. So, uh, that is also something to think about maybe as, like, step two is where you wanna build your company. Um, do you wanna do it kind of like an all-remote company or maybe choose a geography that doesn't have as many interesting companies to compete with? Or do you wanna just go to Silicon Valley and, um, and just have, like, a lot of talent, but be maybe in a very competitive, uh, talent market there?

    4. AGAakash Gupta25:42

       Yeah. And did you guys generally work in person throughout?

    5. MAMatt Arbesfeld25:47

       We... Yeah, this was old school days where, uh, you know, people would... You'd have a dentist appointment or whatever and you'd work from home, but no one would even think like, "Oh, I'm gonna be a remote employee." That wasn't... I, I don't remember that at all being a thing in like 2016, 2017 back in the day. Um, so yeah, we were all in person up until 2020 when we went more of a hybrid approach. And so now we have, like, folks who love to come into the office and we have folks who are fully remote or, or somewhere in between.

    6. AGAakash Gupta26:15

       Got it. All right, so our next step is hire executives. Talk to us about how you built out your leadership team.

    7. MAMatt Arbesfeld26:21

       Yeah. So we, probably around 20 people, maybe first, first 20 people. I wasn't managing everyone, but we had... You know, most people were just kind of individual contributors. They build stuff. Um, and then at some point you realize there's, like, so many important priorities. I can't be working on 10 thing... Like, pushing 10 things forward at once, and you wanna be able to hand off, like, these three problems to someone and these three problems to someone else. And so that's when we first started hiring our, you know, executives. And so, um, I think one of the early ones was a promotion from within. Our engineering leader, Pascal, you know, was a, one of the original engineers and, and then led the team. And then our first external hire was our VP of marketing who, um, you know, we, we sort of had this blog that was working, but we really wanted to scale and grow that and also build some new channels beyond just that blog in case we sort of reached a limit there. And so we spent a, a lot of time, yeah, searching for, for folks and interviewing and, and you definitely will make some mistakes along the way as you're, as you're hiring these executives. But, um, it's important to find people who think you match more so than their, like, hard skills or what they've done. Do you work with them culturally? 'Cause these are your partners. You know, it's almost like you're, you're marrying... It should be almost like a marriage if it works well. Like, you should really be able to work with them and grow and challenge each other to become better. And, and so if someone just has, like, completely different values and mindset towards things, I think that's where you really get a lot of friction in your hires versus someone who you share values and a mindset. You can develop skills and, and work with them through things, but it's very hard if they just have a, a whole different approach to doing stuff than you do.

    8. AGAakash Gupta28:10

       Did you go through some executives who did have a different approach and learn that hard lesson? It sounds like you may have. [chuckles]

    9. MAMatt Arbesfeld28:16

       Yeah, yeah. You know, and p- everyone has been great. I, I don't wanna, you know, dig on anyone, but de- definitely I think there've been folks who... And maybe come from a completely different culture than w- what we were building, and they wanted to take their culture and kind of impress it on the people already there, which I think can be very challenging. You end up having to turn over the whole team, and when you do that, you lose a lot of the knowledge that was built in that function. And so it's generally very difficult if someone comes in and wants to replace 60% of the people. I think it can make you take a step back at least for, for a while, and so you're really betting that whatever new culture they bring in will work better. And I just didn't, you know, haven't found that to be as true as if you have someone that's working, bring someone in that kind of augments and takes the culture to the next level is better than someone who just has a whole different mindset towards what you're doing.

    10. AGAakash Gupta29:10

        On the whole though, you were talking to me the last time we chatted about how you actually have tried to really not turn over the executive team too much and really grow from within.

    11. MAMatt Arbesfeld29:21

        Yeah. You know, I think there's a... There was always this mindset, I think, in company building, like fire fast, be like really ruthless. And I think I've just found that if you hire, if you hire people who care and wanna do a good job, they like to be challenged and shown what they could do better, and they can usually develop whatever skill you notice a lack of or, um... So we've definitely grown a lot with our executive team, and throughout the company I really like to just challenge people, point out where they can grow, and, and I'd say 9 times out of 10 folks rise to the occasion. And if they don't, sometimes they just select out and they'll go interview and, and join somewhere else. But I'd, I'd much rather kind of challenge them, try to help them grow than, um, than just like fire them and try to spend all the time to replace and, and develop someone new who might have the same problems or different problems than that person that you just, you just removed.

    12. AGAakash Gupta30:22

        Yeah. And you tend to lose a lot of cultural knowledge and tend to lose a lot of transition time where people are even performing at whatever lower level than ideal.

    13. AGAakash Gupta30:31

        Hey, let me take a quick break to talk about something that's completely changed my product management workflow: Linear. As a PM, I was drowning in tools, one for planning, another for issue tracking, roadmaps and sheets, and jumping between Slack, Intercom, and AppReviews just to piece together customer feedback. Sound familiar? I was spending more time keeping systems in sync than actually building product. Every time development kicked off, my carefully crafted plans would immediately need updating. I was the human API between all our teams, constantly chasing updates and translating between tools. That's why I love Linear. I can capture customer feedback, shape product ideas collaboratively, quarterback cross-functional teams, and monitor development progress in one place. It cuts through the maze of disconnected systems that were complicating my life. Product teams at OpenAI, Vercel, Brex, and Cash App all use Linear. If you're tired of spending your days keeping different tools and teams in sync, check out Linear at linear.app/partners/aakash. That's linear.app/partners/aakash.

12. 31:23 – 35:35

    Step 10: Raise more money vs bootstrap—competition, venture scale, and regret minimization

    1. AGAakash Gupta31:23

       Now I think we're onto step 10: raise more money and up-level your team. Talk to us about that.

    2. MAMatt Arbesfeld31:30

       Yeah. So in, depending on your space, you might decide to just bootstrap. And, and nowadays I'm hearing you have, like, a one-person company that's 25 million of revenue because all these AI tools. So you might not need to raise any money. But we are- we happen to be in a quite competitive space with analytics and session replay, and competitors spending a lot of money and building huge teams. So we wanted to be able to spend more than our, um, our revenue would, would allow us. And so that's when we s- I think we raised four rounds over the, the course of the company. And in each rounds, we're sharing investors what our plans are, what we project our revenue growth, what we wanna build, and they invest and take a percentage of the company in exchange. And so I think kind of having that cushion of more, more money to, to kind of grow faster than what our revenue would allow just made, allowed us to make some investments ahead of time that we couldn't have made before and, and kind of stay, um, stay with the market and not have our competitors grow faster than us. And then with that money, you also get a chance to launch and build some excitement and hire people and kind of, like, make some noise. And so it also just helps you kind of keep that momentum in the company and, and build some excitement around things.

    3. AGAakash Gupta32:50

       And if you were to do it over again, like now though, in 2024 when more people aren't fundraising, how would you decide, like, "Should I fundraise or should I not?" Because when you did this, I think it was more of, like, a default option, especially in competitive markets, but now I think people have a l- a lot more choice.

    4. MAMatt Arbesfeld33:10

       Yeah. You know, it's tough because most... I think in some ways the markets are even more competitive now. Like, if I look up any A- like, AI category, you immediately find 10 options for it online. And so almost every category is so saturated now that it's hard to, it's hard to win that without having some more money and, and be able to have enough resources to, to hire great people and spend on marketing. So I'd say I would look at the competitiveness, competitiveness of your category, and if it's competitive, it's probably worth raising some money to go after the category. And the last thing you want is to say, "Oh, the thing didn't work because we didn't raise money, and so our competitors outspent us and they grew faster." Like, you don't wanna live with that regret your whole life, so you're probably better raising money if you, if you have a chance of that regret. If you're in a space like, I don't know, like a non... I don't even know any non-competitive markets nowadays, but something where cl- like, no one is really building there, you don't have any competition, that's maybe a time that I would maybe bootstrap more and try to grow organically, um, just because it, you don't, you're not kind of forced to go down this venture path for a business that probably doesn't even have a venture scale kind of revenue potential.

    5. AGAakash Gupta34:30

       Yeah. I think that's one of the biggest ones is what is the scale you want to achieve? Obviously, you wanted to achieve really winning in this market. There's a kind of a different approach out there too, right, where if you just wanna build a one to two million dollar business, it's not venture scale to begin with. Like, venture scale, what are we talking about when we say what is venture scale?

    6. MAMatt Arbesfeld34:50

       Yeah. I guess to go IPO nowadays is, like, 250 million revenue, so yeah, a bit beyond one or two million. Though I even wonder, if you're in a competitive market, it can be even hard to build a two million dollar business because eventually your competitors will build whatever niche you're going after and start to take market share from you. So even... I think it's kind of more based off the market and the potential rather than your ambition as a founder. Like, if you're in a competitive market, you've kind of signed up to compete. If you're not in a competitive market, then you can probably take it, take it easy. But, um, yeah, I just think it's hard to build a small business in a big market.

13. 35:35 – 39:18

    Step 11: Build a second product—AI that ‘watches’ sessions and surfaces issues

    1. AGAakash Gupta35:35

       Yeah. Step 11: build a second product. What was yours?

    2. MAMatt Arbesfeld35:40

       Yeah. I think this is one of the most challenging things for, for product leaders out there is you've built some initial product, people love it, it's growing, you're getting a lot of feedback, but then over time you probably start to see... You start to get actually less and less feedback. People are happy with it. You get on a call, they're like, "Your product's amazing. I have no feedback for you." And you're like, "Okay, what do I do nowadays? Like I'm, you know, I built a perfect product." I guess you could just stop and, like, enjoy, enjoy life. But, um, I think as, like, humans, we're just, our nature is just to keep building and trying stuff. So that's the point when you're thinking, "What's the next big thing that I can build that maybe people are not asking me for directly, but is adjacent and helps complement and build a, a broader portfolio?" And so-Um, choosing that second product, I think, is one of the, the most challenging things as a company, and, uh, doing it right can be the difference between, like, continuing your growth story versus kind of flattening out with that initial product offering.

    3. AGAakash Gupta36:44

       And how did you guys choose your second product, and what was it?

    4. MAMatt Arbesfeld36:48

       So, yeah, given this was our first time, I think it was a bit more haphazard than it probably should have been. But we saw, um... So our core of our product, we capture these sessions of users using our product. So imagine you're a big company, you have millions of these sessions. There's tons of interesting data there about your customers, but you'd have to go through and watch sessions one by one by one, super time-consuming task. And so our second product was what if, you know, what if you could have someone watch all those sessions and report what are the problems that are affecting customers? Where is there friction in the product? And so that was our second product is AI that would basically watch through all these sessions and, and show you issues and errors that were affecting your customers.

    5. AGAakash Gupta37:37

       Got it. So that seems like it's the same user, um, same buyer, like additional use case. Is that the way to think about it?

    6. MAMatt Arbesfeld37:48

       E-exactly. And, and I don't think we necessarily thought about it in such, uh, in such a structured way, but in retrospect, kind of finding people who are already using your product, who love it, who just wanna buy more, much easier than going after a whole another persona that might also be able to use your product. So for us it was, yeah, continued to be engineers and product managers coming in and, and using that new offering.

    7. AGAakash Gupta38:13

       What was the most difficult part of launching a second product?

    8. MAMatt Arbesfeld38:18

       Yeah. I'd say, so as a team, you're so used to people using your product and giving you tons of feedback right away, and anytime there's a bug, they tell you about it. The second product, almost by definition, it has zero users to start. So, um, any... You're like, the team is so used to getting this constant feedback, and now you have a- you look at the graphs of usage of that second product and it's like, "Wait, no one's using it. Why am I working on this thing?" And you start to kind of get a little discouraged about the product. And so I think there's this gap as you're building new products where people are not using it as much, but you have to just believe in the, in the mission and vision, and know that if you build it and keep iterating, they will start using it and, and you have to keep making it better. So I'd say that was the, the most challenging part is that kind of like disillusionment as you launch your product and, like, usage is not as high as you think and, and just takes years to really get that product off the ground is kind of what we've learned over the years.

14. 39:18 – 44:10

    Step 12–13: Diversify channels and broaden the suite—golden eras, new GTM bets, more products

    1. AGAakash Gupta39:18

       All right. We are on, if I'm counting correctly, step 12, expand your go-to-market channels. So how did you guys scale your distribution?

    2. MAMatt Arbesfeld39:28

       Yeah. So we were kind of forced to in some ways. So we had this blog. People would search on Google, "How do I do something?" No one is really searching on Google anymore. Everyone's going to ChatGPT to, to do that. So we, we sort of saw this happening a few years ago, where we realized the blog couldn't grow forever. We had to diversify and find new ways to, to go to market. And so over the years we've, we've done a lot of new things. So we do a lot of trade shows now, so events where product folks are there. We do a lot more on LinkedIn now and YouTube and, and places beyond just the Google, Google blog. So I'd say we started by kind of placing five, six, seven different bets of things we think might work, and we saw which ones took off, which were having traction, and then we directed more resources towards that problem, um, as we were starting to see them take off, and then we would sort of shutter the, the investments that weren't paying off after, say, six months.

    3. AGAakash Gupta40:27

       And I think one thing that you did though really wisely is when you really understood Google, you actually invested in it. And I think last time we talked you said that you might have even been able to go more intensely on what's working when it's working.

    4. MAMatt Arbesfeld40:39

       100%. Yeah. When you have that first go-to-market channel that, that works and you're in a competitive space, I think you have to pour the gasoline on as, as fast as you can. And I think you're always, especially as a first-time entrepreneur, you kind of wanna be a little cautious and not spend a ton of money and, and kind of grow sustainably. But when you have something that's working in a go-to-market channel, like you're kind of in a golden era, right? Like, if you think of the, the Roman civili- like, when you're in a golden era, that's when you get all your gains. And when you identify that golden era, you have to just go all in. And so, um, that's true, kind of identifying that golden era and, and, and doubling down, I think, is, is something I wish we had done a little more, uh, when we had that initial traction and, and so something we're always excited for when we get that kind of, uh, go-to-market channel that's working.

    5. AGAakash Gupta41:30

       Yeah. I was just recording a podcast with someone who was looking at my top performing LinkedIn posts, and I was living in a golden era in 2021 when I could publish every post getting a million impressions. We do not live in that era anymore, so how I wish I hadn't taken a three-month break in that year. I think we can all relate to that one. All right, our next step, build more products.

    6. MAMatt Arbesfeld41:53

       Yeah. So I'd say nowadays, I, I... So back in the day, the conventional wisdom was you build one product, then you build your second product, then you build your third product. Nowadays, you look at some of the companies out there, like a Rippling, I think they offer 20 products, right? Or Ramp, they offer 15 products. Maybe this is combination of all the AI tools and engineers are just more knowledgeable and they have more product skills, but you- these spaces are getting so much wider and more distributed, and so it's important to build a third and fourth and fifth offering probably ahead of when you, you think you might otherwise have to. And so-For us, that was we went after product analytics as well, so building, um, funnels analysis and heat maps and journey analytics on top of the data we were already collecting. But kind of identifying what are the third, fourth, fifth products in your space, and maybe having an engineer or two go after those spaces and, and build out MVPs would be my recommendation to kind of a scaling company.

    7. AGAakash Gupta42:56

       Yeah. I think that there's probably a certain ARR slowdown, right, that you experience on your core when you wanna layer in new ones, or is it... How do you think about, like, the right indicators, I guess, to really understand, "We had these two products. They're doing well. Let's go ahead and build three, four, five, and six"?

    8. MAMatt Arbesfeld43:17

       Yeah, it's hard. I feel like there's no hard and fast rule. Um, one question is, what do your engineers... What, what's kind of your engineering team skill set? And if you have folks that love to build zero to one products, they might be better, um, used to go after those new categories rather than just kind of iterate on an existing product area. You also wanna look at, for your existing product areas, how much more is there to build on top of what you're already doing? Um, so like for us, there's so much more we're doing with AI on our existing categories, so we have a lot of investment there. But then y- you kind of have to balance that against where's the market going and, and kind of what needs to be the breadth of my product suite. So I'd say, yeah, there's no hard and fast rule, but it's kind of deciding what is the best use of your resources to, uh, to help your customers.

15. 44:10 – 45:44

    AI product realities: measuring impact and the accuracy/trust problem

    1. AGAakash Gupta44:10

       And how are you guys handling that now, your AI investments? How are you tracking, measuring, organizing those?

    2. MAMatt Arbesfeld44:18

       I, yeah, it's still... I think everyone's kind of learning in this, in this space. I'd say the, the best thing to do is always just talk to your customers. Um, like my favorite thing is when I get on a call and they're like, "Oh, y- I just used your AI. It's incredible." Like that just sh- No one would say that unless they actually like what you're doing, and so that's the best validation. Um, that and spending money on, on your thing, on, on the AI. Um, before that, use- looking at usage, looking at session replays of people using the AI features, kind of seeing what engagement looks like with, with your AI is really good, but there's nothing better than people spending money and just telling you how much they love it.

    3. AGAakash Gupta44:56

       What's been the biggest sort of stumble you guys have had building AI products?

    4. MAMatt Arbesfeld45:03

       Yeah, I'd say the hardest part of any A- AI product is the accuracy and quality. Like, as anyone knows who's used an AI or built an AI product, it's really easy to get an MVP off the ground. But then as you keep using the thing, you're like, "Wait, that's wrong. Like, that's incorrect." Like, and you start to lose trust in the data. Like, you ask ChatGPT a question, it will just, if it makes something up, you then kind of question what it's doing for the next few months. Like, you don't really believe it un- unless it cites something for you. So building very accurate AI products and ensuring that accuracy before you launch it too widely, I'd say is the biggest challenge we've run into with AI.

16. 45:44 – 53:43

    Step 14–15 and founder reflections: partnerships, moats, ambition, and the CEO grind

    1. AGAakash Gupta45:44

       All right. Step 14, second to last one, develop partnerships. Talk about that.

    2. MAMatt Arbesfeld45:49

       Yeah. So I think a lot of people, when they start a product, they're like, "You're so tiny, and you're against these huge compa- You know, huge companies might come into your space," and you're worried about how do you have a defensive product, or what's your defensiveness? So what would stop someone from copying you and just cloning your product? And so that's sort of always true throughout the lifetime of the company is, like, someone could always come and clone what you're doing and compete. Google could come in and, and just build a replica, or Microsoft or Amazon. So the only ways really to hedge against that are to get out in the ecosystem and have people know about you and start to connect all your data together. So if you think of something like Salesforce, it's so widely connected to all these different products in the world that it's really hard for someone to build a new Salesforce. You have maybe only one or two companies, like a HubSpot, that have been able to compete. And so the more you can develop an ecosystem and partnerships or, or kind of marketplace around what you do, the more sort of, uh, defensiveness and, and moat you have around what you're doing.

    3. AGAakash Gupta46:57

       And how have you guys approached that?

    4. MAMatt Arbesfeld47:00

       Yeah. So we've built a lot of integrations with various products in the ecosystem. Um, that's been a big, big thing. We're continuing to work with, with partners who are in kind of adjacent categories, so they will recommend and build connections with us. So kind of seeing who are our adjacent people in our ecosystem and developing partnerships and integrations with those kind of companies has been kind of key to our strategy.

    5. AGAakash Gupta47:25

       What have been one of your most important partnerships?

    6. MAMatt Arbesfeld47:29

       Yeah. I'd say we, we do a lot with Expo, who's a, a mobile app development framework. Um, so anyone who's building Expo can really easily get access to, to LogRocket in their mobile app. That's been a really great partnership for us. We also do a lot with the, um, help desk tools, like Intercom and Zendesk, so that we, we wanna... When someone has a problem using your software, you really wanna quickly see what were they doing in your product. And so having a deep integration with those help desk systems has been really important.

    7. AGAakash Gupta48:04

       Very cool. What's our last step?

    8. MAMatt Arbesfeld48:07

       Take over the world. Uh, w- we're not quite there yet, uh, so I guess it, uh, it's a fu- it's a future step for us. But, um, yeah, I think the, the goal is always you want everyone in the world using your product, right? That's, that's what we're here to do is, um, if you build a great product, you kind of regret when someone doesn't use it and doesn't get a chance to, to, to try it. And so we're always... You know, our goal is to be on every website in the world and help everyone make their perfect software, uh, with what they're doing. So, um-Definitely a, a future goal, but, um, yeah, hopefully not too far in the distant future for us.

    9. AGAakash Gupta48:46

       What has been the lowest point in your journey?

    10. MAMatt Arbesfeld48:51

        Hmm. The lowest point. You know, I'd say one of the tough things, like, as a, as a CEO, you see all the bad things in the company, right? Like, if something's good, people will often tell you about it, but if something's really bad, they definitely tell you about it, 'cause they don't want you to... they don't want you to be surprised later. So almost every Slack I get is a problem that someone needs my help with, or else they would've solved it already. So I'd say there hasn't been, like, a global low point beyond, like, when you're... There's just a constant stress of starting a company 'cause you're just bombarded with problems every single day. Um, so I think the hardest parts are, the hardest times are just the ongoing grind of dealing with problems day in and day out, um, rather than any kind of specific low point that I can, I can think of.

    11. AGAakash Gupta49:46

        If you were starting today and you worked through our first few steps, you started working on projects, you built some skills, you understand a problem domain, how would you approach building an MVP? Would you go to market really fast with something like a Lovable and just, like, upgrade it a m- little bit with a cursor and just get it out there? Would you work with some early alpha/beta customers and get it really hardened before you do a big launch? How would you handle that?

    12. MAMatt Arbesfeld50:12

        I think right now there's a big opportunity for AI-enabled services. So if you think of what AI will enable, it's automating a lot of work that humans are doing, but it's not quite there yet. So there's this gap where people wanna use AI for something, but then when you try to use the AI, it ends up wrong and hallucinating, and so you n- so I would probably start identifying one of those categories, seeing... building a, a basic UI, maybe with a Lovable or, or Bolts, but then actually powering the service just through my own human effort. Um, so almost, like, making it feel like an AI to the customer, but then you can ensure the quality is really high by providing your services on top of that, top of that AI. I think there's a big opportunity to build products in that, in that space.

    13. AGAakash Gupta51:03

        Awesome. Now that you guys are more successful and you're not on the ramen budget anymore, how [chuckles] does that feel to, like, be leading a successful company?

    14. MAMatt Arbesfeld51:15

        It honestly is more stress. [chuckles] I think, uh, i- in some ways, it's very simple when it's, like, two or three of you and there's, like, no stakes. You know, now there's like, uh, you know, like I said, there's always some issue that needs solving and, and some, um, something that you're, you're fighting. And so, you know, I'd say definitely the stress levels get higher as, as the stakes, as the stakes grow. Um, but at the same time, you know, it's very fun. You just have those moments when you, like, you hit the number for the quarter, or you hire a great person, or the team's all together and celebrating. Like, those are even more joyous. So I feel like you have higher highs and lower lows as a company goes on and, and I just kind of try to relish the highs as much as possible now.

    15. AGAakash Gupta51:58

        And now that you are where you are, what are your best sources of information? Are you talking to a lot of other CEO founders? Are you talking to your investor and board members a lot? Are you reading content online? What is the places where you're finding stuff that, "Okay, this is helping me in my role now"?

    16. MAMatt Arbesfeld52:15

        So in terms of product, nothing's better than customers, especially seeing customers in person and hearing them and seeing what they're doing and their problems. That is the im- really important for product. In terms of being a founder, CEO, um, I'd say definitely talking to other CEOs, talking to your team, getting their feedback, reflecting on what's going on in your mistakes and your learnings from it, I think has been big, and just talking through those things with, with my co-founder and others, um, I think have all been kind of sources of learning in terms of how to run the company.

    17. AGAakash Gupta52:53

        Awesome. Well, I have learned so much. This has been super fun walking through 15 steps. I hope that this helps you as well. If people wanna find you online, um, where can they do that to learn more?

    18. MAMatt Arbesfeld53:06

        Yeah. So I'm LinkedIn, uh, Matt Arbesfeld, and then definitely come check us out at logrocket.com.

    19. AGAakash Gupta53:12

        Matt, thank you so much.

    20. MAMatt Arbesfeld53:14

        Thanks, Aakash. Good to talk.

    21. AGAakash Gupta53:16

        All right. Perfect. I really hope you guys enjoyed that episode. It would mean a ton to me and the team if you could please subscribe on YouTube, follow on Apple and Spotify podcasts, and leave a rating and review. Those ratings and reviews really help grow the show and help other people discover the show, and they help fund the production so that we can do bigger and better productions. Can't wait to share the next episode with you. Until then, see you later

Episode duration: 53:44