He killed a viral app with 50k users. 2 years later, he hit $10M ARR and raised $30M from Sequoia. | David Paffenholz (Juicebox)

David Paffenholz (00:00:00):
Yeah, I think I had one to two million views on LinkedIn, which is a pretty high view count. The video went viral. We got a lot of people trying it out overnight. We went from zero users, zero customers, no one had ever paid us anything before, to having like a hundred paid users on pretty cheap, fifty-dollar-a-month subscriptions. Almost all of our growth is purely inbound. Word of mouth is by far the biggest channel. Over half of our signups come from word of mouth. We got over 500 free user signups a day. Nowadays we do other things as well, so it’s diversified a bit, but I’d say up to the ten-million ARR mark it was very driven by PLG and then kind of sales-assisted upsell for a larger customer. Especially in the early days in 2024, it was really like marginally better. It was already a big unlock because it meant that this new workflow gave a better output than the existing workflow. And the new workflow had not even been optimized yet. The 10x better always has to be caveated with how important the thing being worked on is, and if it’s a really important thing, then 2x better is a huge outcome for the end user, let alone 10x.

Pablo Srugo (00:01:03):
And you've grown what, like 3x this year?

David Paffenholz (00:01:04):
This year, closer to 10x.

Pablo Srugo (00:01:05):
Oh, wow. Okay, so you've crossed 10 million, top line.

David Paffenholz (00:01:08):
Oh yeah, we announced, so 10 million we announced as part of our Series A announcement.

Pablo Srugo (00:01:11):
That's an insane year.

Previous Guests (00:01:15):
That's product market fit. Product market fit. Product market fit. I called it the product market fit question. Product market fit. Product market fit. Product market fit. Product market fit. I mean, the name of the show is product market fit.

Pablo Srugo (00:01:27):
Do you think the product market fit show has product market fit? Cause if you do, then there's something you just have to do. You have to pick up your phone. You have to leave the show five stars. It lets us reach more founders and it lets us get better guests. Thank you. David, welcome to the show, man.

David Paffenholz (00:01:44):
Thanks for having me on.

Pablo Srugo (00:01:45):
Dude, I mean, I’m hyped up, man. I’m hyped up. I’m ready to make this happen. You started, like, I mean, I wouldn’t say long ago at all, right? But like 2020, 2021, and seems to really have hit an inflection point over the last year or two. I mean, you’ll tell me, but let’s say a year or two. You just closed a $30 million Series A by probably the most historic, let’s say, firm in the world, by Sequoia. Just this summer it was announced in September. So looks from the outside like things are off to the races, but excited to hear what it’s really like inside.

David Paffenholz (00:02:16):
Yeah. Excited to chat. As you hinted at, it took a while for things to work.

Pablo Srugo (00:02:20):
Let's start with the beginning, man. I mean, maybe just like, what does Juicebox do?

David Paffenholz (00:02:23):
So Juicebox is an AI recruiting platform. Really what we do is help companies find the very best talent for their teams. And we do that using natural language search. So you can just describe who you’re looking for using a short prompt, and then you’ll start seeing top matches for the role. From there, we help you reach out to them, manage those contacts, and more. So kind of all in one top of funnel for recruiting.

Pablo Srugo (00:02:45):
Is it kind of like the next level of like, you know, you think like LinkedIn recruiting is kind of like a core outbound piece. Is it like just the AI version of that?

David Paffenholz (00:02:53):
Yeah, that's a good way to think about it.

Pablo Srugo (00:02:55):
And when did you start this business?

David Paffenholz (00:02:56):
Uh, we started the company in 2022 and then we pivoted into the current shape of the business in a mid 2023.

Pablo Srugo (00:03:02):
Okay. But you incorporated 2022.

David Paffenholz (00:03:03):
That's right.

Pablo Srugo (00:03:04):
Okay, cool. So maybe walk us through, give us a little bit of your background and then let's go deeper on like kind of the post COVID, like what were you doing from 2020 till 2022? Just cause that time before is always really relevant.

David Paffenholz (00:03:15):
For sure. So I’m originally from Germany. I grew up there. I came to the US for college, studied econ, but almost immediately fell into tech. I did a couple of internships in companies and also on the venture side. And then I spent a year during COVID, which I took off from college, working at Snap. Initially as an intern, I was there for three months at first and then ended up staying for almost a year. It was a really fun experience. I learned a ton. And I also got to know my co-founder during that time.

Pablo Srugo (00:03:41):
What year was that that you were at Snap?

David Paffenholz (00:03:43):
Yeah, mid 2020 through mid 2021. And so college had, like, I think COVID started breaking out in March or something, and then by summer I was interning at Snap, and then I ended up just continuing that for a full year.

Pablo Srugo (00:03:56):
Snap is one of those companies. I mean, obviously, everybody knows about it. But I just wonder what it’s like inside of Snap. On the one hand, it’s such a well-known, well-used product by a certain segment of the population. And then you have the big behemoth that is constantly pounding on you, and then TikTok being a newcomer that really took off. I’m just curious. What’s it like? I mean, first of all, what’s the culture like at Snap? And second, when you’re there, what is the perspective of that competitive dynamic?

David Paffenholz (00:04:24):
Yeah, it was interesting. I was on the international growth team, which basically meant we were focused on what growth we could drive in international markets. And that growth was primarily focused on usage or user growth rather than revenue growth. And so Snap actually makes a decent amount of revenue, I think more than one would maybe assume, and a lot of that is advertising driven by the U.S. and then a bit less the European markets. And, you know, the goal of the team was to show that same usage growth internationally so that in the future those could also be viable markets. And so we thought a lot about what the competitive positioning is, how that changes from market to market, what people actually use, and what Snap does incredibly well, which is they kind of capture this segment of the audience, which is like 13 to, say, 25-year-olds, and almost everyone in certain markets is a daily active user. But then I think it’s a little bit less clear what users do beyond that as they grow older and maybe don’t use the product as much or migrate off to different platforms. And I think that was something we thought a lot about while I was there and also something that maybe is still unsolved.

Pablo Srugo (00:05:25):
Yeah, we had one of the early investors on the PMF show talking about so many innovations that came out of Snap, obviously disappearing messages, but even things like Stories and ordering content in reverse chronological order. There are so many things they started that now, obviously, many others have done. So they’re a very product-forward company. So maybe tell me, what happens after Snap?

David Paffenholz (00:05:45):
Yeah, so while at Snap, I was kind of helping organize this competition with a few friends from college. And the competition was designed for high school seniors. And so it allowed everyone to submit a research paper, really with a technical focus. And it was COVID themed because everything was COVID themed at the time. And so it was about social distance.

Pablo Srugo (00:06:04):
It's like 2021?

David Paffenholz (00:06:05):
Yeah, exactly 2021. It was late 2020. And so on social distancing and the prize was you got to meet Stephen Wolfram. And so I had helped organize this competition and we had ended up having a good number of students participate, I think in around 10,000 or so, roughly that range. And one of them was Ishan, who ended up winning the competition, and I got to know him through that. He actually ended up becoming my co-founder. And so at the time we just got to know each other. We were working on a couple of projects with a few other people as well, and they were really just startup-esque ideas.

Pablo Srugo (00:06:38):
But you're still at Snap at this point?

David Paffenholz (00:06:40):
Yeah, exactly, 2021. It was late 2020. And so on social distancing, and the prize was you got to meet Stephen Wolfram. And so I had helped organize this competition, and we had ended up having a good number of students participate, like I think in the around 10,000 or so, in roughly that range. And one of them was Ishan, who ended up winning the competition, and I got to know him through that. He actually ended up becoming now my co-founder. And so at the time we just got to know each other. We were working on a couple of projects with a few other people as well, and they were really just startup-esque ideas.

Pablo Srugo (00:08:03):
I think it's also the maybe the context, right? Like it's this kind of that lean back, lean forward kind of thing, right? Like maybe in this context, it's less lean back, lean forward and kind of like, you know, with podcasts, I think a lot of times you're doing something else. Maybe you're driving, you're walking, you're cooking, you're, you know, it's kind of in the background, you're tuned in, but you don't need to be like 100%. Where as with even something like short form, yeah, it's leaned back, but you're watching the thing, you're on the thing, and you need to be constantly, if it's not interesting right now, you flick it, whereas if it's not interesting on a podcast, you maybe just tune out for five minutes, and you tune back in, so I wonder if that also plays a role, like you said, for short form, the stimulation level needs to be so high, because otherwise, you won't just go to the next clip, at some point you'll just get out of the app, and do something else.

David Paffenholz (00:08:45):
It's a good point. It's like if you're visually looking at the app, it needs to capture your full attention versus as you point out on like a podcast, it doesn't necessarily need to capture your full attention. Just enough to keep your audio attention, I guess.

Pablo Srugo (00:08:56):
That's right.

David Paffenholz (00:08:57):
Yeah. And so that was our experience there. It was fun to build, but it was also kind of rough because it never, you know, we got like a few hundred users on there, but no one was retentive. And I guess all the classic initial failed college social media startup idea. The best thing about it though was that Ishan and I really enjoyed working together. And so we gave it another shot shortly thereafter with a music-focused app. And so similar concept, still short form, but in this case you could connect with your Spotify and then get personalized music recommendations. And there was a social element to it as well of sharing different song snippets. And that idea worked better. We got 50,000 users on there pretty quickly. We had some TikTok virality of people posting about us and more. And it was kind of the first feeling of, oh, this is a product that people actually want to use. And this is something that could work. And still very much from a perspective of thinking of, you know, we want to build a fun product.

Pablo Srugo (00:09:49):
This was before like Spotify Discover or whatever, like they were doing less, I think, with playlists then?

David Paffenholz (00:09:54):
Yeah, exactly. I was on the international growth team, which basically meant we were focused on what growth we could drive in international markets. And that growth was primarily focused on usage or user growth rather than revenue growth. Snap actually makes a decent amount of revenue, I think more than one would assume, and a lot of that is advertising driven by the U.S. and then a bit less by the European markets. The goal of the team was to show that same usage growth internationally so that in the future those could also be viable markets. We thought a lot about what the competitive positioning was, how that changed from market to market, and what people actually used. What Snap does incredibly well is capture this segment of the audience that is 13 to 25 years old, and almost everyone in certain markets is a daily active user. But then I think it’s a little less clear what users do beyond that as they grow older and maybe don’t use the product as much or migrate to different platforms. That was something we thought a lot about while I was there and also something that maybe is still unsolved.

Pablo Srugo (00:10:52):
Yeah, at that point, what stage was the product? Where were you at when you got into YC?

David Paffenholz (00:10:56):
Yeah, when we applied to YC, we had built the music app, but we knew we didn't want to continue working on that. And so our written application was very heavily about the music app. But by the time we got to the interview stage, we were, hey, you know, we wrote a lot about the music app thing, but we're not sure if this is going to work. And so we had a few other ideas as well. And that was largely around the recruiting space, but no real product or no traction.

Pablo Srugo (00:11:20):
But you pitched that, because I know the interviews in YC are very short because it's high volume, so if you don't have something crisp, I don't know how you make it in sort of thing.

David Paffenholz (00:11:28):
Yup, so we had something crisp on music, or we had an idea around it, but we also knew if we got that question, they sometimes ask the question, oh, do you have other ideas, or are you thinking about doing something else, basically when they're skeptical about what you're pitching.

Pablo Srugo (00:11:41):
Right.

David Paffenholz (00:11:42):
We also knew we were, you know, we were not married to that idea. And we indicated that in the interview too, that, you know, we're open to exploring something else. And so that ended up being important because we did that basically right away. And, you know, we decided, hey, the music thing is probably not going to work. And so let's start fresh and see where we can find something that we really believe in.

Pablo Srugo (00:12:00):
Why do you think, maybe this is a better question for YC, but why do you think you got in? I mean, so many people tried to get into YC, and I'm sure they come up with very specific pitch, specific market, all this stuff, I'm all in. You seem to, I don't know, you were casual about it. You got something that was working, but you're also, ah, but I might do these other things. And yet they saw something in you and wanted you in the program. Why do you think that is?

David Paffenholz (00:12:19):
Yeah, I think we did it. We showed that we could execute, the product had real usage on it. It had over fifty thousand users. That’s not that easy to get.

Pablo Srugo (00:12:27):
Yeah, it's true.

David Paffenholz (00:12:28):
And it had some virality to it as well, we built something that clearly worked. And then I think the other thing that stood out is that Ishana and I had built two things together at this point, and both of them didn't really work, but we still wanted to continue building and do something. And so I think that co-founder dynamic was also really important. Yeah, so I'd sum it up as the combination of, you know, we built something that worked, we knew we were capable of building something that would work, and then two, us as co-founders did that together. And so I think that was the bet on why we might be able to succeed in YC.

Pablo Srugo (00:12:55):
How quickly into YC do you decide that you don't want to go down this music app idea?

David Paffenholz (00:13:01):
Before it started, literally maybe a week or two after we got in. So it was super quick. We were even embarrassed. We were like, oh, you know, how do we approach this with our YC partner? You know, they're going to.

Pablo Srugo (00:13:11):
As though they hadn't seen that a million times by that point, right?

David Paffenholz (00:13:14):
We were really worked up about it. In reality, I think, I don't even know if he remembers this, but we were very, very worked up about it.

Pablo Srugo (00:13:24):
What was the key catalyst to making that? It's not an easy decision. You have fifty thousand users, things are kind of working. A lot of people would just say, let's just try to get to a hundred, let's just hustle or whatever. What was the catalyst for you guys to decide, no, this is not the thing?

David Paffenholz (00:13:37):
Yeah. So two things: one, retention numbers. And that was the easiest metric to look at, and I had seen what those metrics looked like at Snap, and they were phenomenal there. And then I saw what they looked like for us, and our day thirty user retention was catastrophic. People would not come back, not even after seven days, let alone thirty days. And the only thing that was growing for us was the top of the funnel. We had new users downloading the music app because it was, I think, catchy and had some virality to it, but it didn't have a habit. It didn't have something that people would be coming back to every day. And so the number of retentive users was just very low.

Pablo Srugo (00:14:13):
I talk about this in many different vectors. When it comes to team, I would say especially, but in your case, in terms of metrics and retention specifically, it's interesting how important it is to have seen what excellence looks like. Obviously, you can look up benchmarks and see some of this stuff, but I find unless you've been there and experienced it, you might not believe it or maybe fall into, oh, this is the average, so we're doing better than average. But when you see what true greatness is, it's kind of hard to settle for anything much lower than that.

David Paffenholz (00:14:40):
Yeah. Yeah. It's a good way of describing it. Even today, with today's product, we have this revenue analytics platform that we use, and they actually benchmark. Here's what tenth percentile growth looks like, ninetieth percentile growth. Then they plot that all on a chart and put our revenue number on that chart as well. And so it's actually very interesting to see what best in class means and where we stand on that.

Pablo Srugo (00:15:02):
How does recruiting come into the picture?

David Paffenholz (00:15:05):
Yeah. So, you know, we were doing YC, and we knew we needed to find something new. A lot of other people in the batch at this point were also pivoting. A few weeks into the batch, you start seeing the first other companies switching ideas. And we started seeing other companies do this very quickly, within two days or something, they had a new idea and they were going for it. We struggled to do that. It ended up taking us multiple months to figure out what idea we wanted to double down on. In fact, more than that, five or six months total.

Pablo Srugo (00:15:34):
So that's post-YC at that point?

David Paffenholz (00:15:37):
Yeah, during YC and post YC, we had iterations on the idea of recruiting. We knew we wanted to solve something in recruiting from the employer perspective, or at least monetize that way. And we wanted to go after a big problem, but we didn't exactly know what that meant. We had this overarching thesis that there's an opportunity to do better matching between opportunities and people working in those opportunities, but we didn't know what that exactly would look like as a product.

Pablo Srugo (00:16:02):
Where did that come from? For you to spend five months trying to figure things out means there was some pretty powerful driver to take you that way. Once you say no to music, you could do anything. It's like the world is yours. So you decide to go down recruiting, and you take a few months to figure out this is going to be a thing. What is pulling you in that direction to begin with? Was it a personal experience with something else you were seeing?

David Paffenholz (00:16:24):
So both my co-founder Ishan and I had worked essentially all throughout college, and in his case, even high school. All the opportunities we had were ones that we cold outbounded ourselves, and we created those opportunities. As an example, the work I did at Snap came from a college consulting group case that we did for them, and I was able to convert that into first an internship and then staying there long term. I also did some part-time work at different VC firms during that time, also by cold outbounding, trying to create opportunities for myself. It was very high friction, ninety-nine percent of those emails never got a response or never led to anything, but it was possible to create those opportunities, and there was a real need for the company that was being fulfilled too. If the kind of way of creating jobs or opportunities was that high friction or required that much effort, then clearly something was not going well in the labor market overall, or there was something missing if that was the best way of finding an opportunity. That was the overarching thesis that we both had, and we believed quite strongly in. It was also one that personally resonated with us, where most B2B software requires having worked in a role where that pain point was felt or experienced. But we felt we experienced it from the candidate's perspective, and we knew there was something that we could build.

Pablo Srugo (00:17:40):
There's been a lot, and I'm sure when you started to talk, especially within YC, there are people there that know every startup that tried anything. But when you think about recruiting, there have been a lot of companies that have tried to solve that problem. LinkedIn is the big one that is solving that problem, obviously building a huge business around just candidate profiles and ultimately recruiting. I remember Hire.com, it was really hot until it wasn’t  and there was a bunch. What were some of the things? Take me deep into that five-month journey, because finding the right problem and solution is a huge piece of ultimately getting to product-market fit. You had the problem. It was a problem that many others before you had seen. What are some of the key moments in that five-month journey that made you decide where to end up and how to really solve it?

David Paffenholz (00:18:23):
Yeah, there were two key learnings we had in that five-month journey, and both of them were really heavily influenced by the recruiters we actually got to know during that time. And so the first learning we had to get to was profile search or profile review. Basically, something where it involves looking at a profile, be that a LinkedIn profile, a resume, or any other type of visual version of a profile, is a very substantial percentage of recruiter time. Let's ballpark it at half of recruiter time. And that really came from just spending time shadowing recruiters on literal Zoom calls. And so we would just see what people would be doing, and they'd generously walk us through their workflows, and we'd see the amount of time spent on some type of profile review. It was clear that that was such a core part of the job, but it was also not clear what a solution to that could look like. And LinkedIn has a really big ML team, and they've done a ton on search ranking, and all of that goes back to that idea of the search problem. Can you surface the right person at the right time? And then the second kind of big learning or big thing that happened, and this was really just by coincidence of timing, is that ChatGPT launched in October two thousand twenty-two. And so suddenly we had this kind of form of intelligence alongside the GPT three point five API launched as well. And so we had this kind of intelligence available by API, which could actually solve that profile review or profile search problem. Because if broken down into small enough steps, we could look at the different things that one assesses for in their head. So, let's say there are five different criteria we're looking for for a given role, and we kind of mentally assess a profile against those, we could just have an LLM do the same thing. Even when viewed on an individual profile basis, it might be perceived as assessment, versus if you do that at really large scale, that's actually search. By reviewing tens of thousands of profiles, you're searching through those profiles. And so that was the big technology unlock we had at the time. By November two thousand twenty-two, it was clear to us, okay, we can bring those two things together. And we started iterating on what that product could look like. And so that was the main learning journey for us or what got us to that point.

Pablo Srugo (00:20:22):
I know you had this one insane, viral post that you made in the very early days on LinkedIn. I think seven thousand likes or something like that. I'm not sure how many views, you can tell me. Maybe tell me more about that. And was that kind of the first product that you put out?

David Paffenholz (00:20:34):
Yeah, so we launched then in May two thousand twenty-three. PeopleGPT by Juicebox, name very much chosen to try to drive some virality at the time because everyone was familiar with ChatGPT. The concept of LLMs was still pretty new to most people, like the average person has probably not heard of them. But a lot of people had heard of ChatGPT at the time. And so we launched PeopleGPT. It was a ninety-second video on LinkedIn.

Pablo Srugo (00:20:58):
When was this, by the way?

David Paffenholz (00:21:00):
May 2023.

Pablo Srugo (00:21:01):
Okay, so it wasn't like right after? Six months after, but still very much in that early wave.

David Paffenholz (00:21:06):
It took us a while to get there and to build because by the time we launched that, we actually had a self-serve version live. So you could actually use the product, and you could pay for it. And so in that ninety-second video, we just kind of walked through what the product did. At the time, it was pretty simple. It let you do an initial search. So you could describe what candidate you were looking for. It would return a list of profiles. And then it had this chat interface to ask questions about them or kind of go in more depth on your search. The video did very well because I think it captured what people in our customer audience, which was recruiters or maybe even people hiring more broadly, were looking for bringing that innovation of AI or LLMs into the recruiting sector. And so that was the broad message market fit that we had.

Pablo Srugo (00:21:49):
How many views did it get? Do you remember?

David Paffenholz (00:21:51):
Yeah, I think one to two million views on LinkedIn, which is a pretty high view count for a piece of content there. And it had fairly large engagement. I mentioned roughly seven thousand likes.

Pablo Srugo (00:22:01):
And it wasn't, by the way, what I was expecting when I watched it, like a Cluely type of hype marketing video, stuff you see these days. I mean, it wasn't that common back then. And this was not that. It was very much like a classic: Here's my product. Here's the demo. Here's what it does. And yet clearly there was something that resonated because, like you said, it totally took off.

David Paffenholz (00:22:19):
Yeah, it was literally like a Loom recording. The editing that I think we did well is that it was very fast paced. So pretty quickly, you could kind of see what the flow was. But yeah, it was not very sophisticated at all—filmed in a WeWork with Loom. And so the video went viral. We got a lot of people trying it out. I think overnight we went from zero users, zero customers, no one had ever paid us anything before, to having one hundred paid users on pretty cheap, fifty-dollar-a-month subscriptions.

Pablo Srugo (00:22:46):
So for us, we were like, okay, if this goes well, we'll get five paid users from this. We'll get feedback, and we'll continue to improve. And so that was kind of our, I think our benchmark. You know, obviously we were hoping for the best, and we ended up getting a lot more usage than we expected. And I think that was really big too because it meant we had momentum, and we were convinced we were going in the right direction.

Pablo Srugo (00:23:11):
Right.

David Paffenholz (00:23:11):
And so that was kind of the success moment, and then it kind of continued for ten days or so afterwards, where there were still new people finding out about it, and the virality kind of continued there. But then, shortly thereafter, we kind of started realizing, hey, these people are all turning, or we're not really solving their problems. Like, we're not actually helping them find new candidates, or only in very rare cases was it successful. That was good feedback, but we also really just took it as feedback for us. We felt like we had validated the message market fit, or that there was something we could build here, or something where we should continue building, and there would be people who are interested in trying it out. Now we have to figure out that the product actually does what we say it does to a sufficiently high level of quality that people want to continue using it.

Pablo Srugo (00:23:52):
So I understand high level, recruiters, whether they're in-house or, you know, offshore or outsourced recruiters, they spend a lot of time looking for candidates, trying to fill roles. That makes sense. Analyzing resumes to see whether they fit certain roles and going outbound, that totally makes sense. I guess my question is, there are platforms to do that stuff. There is LinkedIn or Guru being the main things that probably everybody has in their tool set. What was it about what you were even pitching then, about your message, that still drew people versus them saying, I could do that over here as well, maybe not as slick? What's really the difference? What was it that you were selling specifically at that point that seemed to resonate so much?

David Paffenholz (00:24:35):
Yeah, I think it was the notion, and this is still true today, that the majority of recruiting software is really dependent on two things. One, the recruiter actually using the software, and two, the software itself. And so oftentimes the software itself is built in a way that it's safe and comprehensive. You can just click through profile by profile, and nothing can really go wrong, but also nothing can really go right. There's no real value add from the software. And when we built Juicebox, the concept was, hey, we're actually going to try to save some of the time that the recruiter spends using the software on top of providing the actual software. So our goal was always to do the work for the user rather than provide a platform for the user to do the work, if that makes sense.

Pablo Srugo (00:25:18):
And does "do the work" mean what? Like setting filters? Like you type in natural language, and it sets filters for you? Is that the main unlock? I know it's evolved a lot since then, but I'm just talking about the beginning.

David Paffenholz (00:25:28):
Yup, yeah. Setting filters was V1, but then, once you've set the filters, your hit rate is probably still only at best twenty to thirty percent of filtered profiles you're actually looking for. And so then the actual work is going through all of those profiles and saying, is this someone that we would actually want to speak to in an interview? Oftentimes, there are a lot of different reasons to object, like, oh, they didn't work at a company previously that we would target hiring from, or they have some skill matches but not all of them a lot of different objections that can exist. And so what Juicebox did is it kind of did that work of reviewing and getting down to the twenty to thirty percent hit rate for you. Ideally, it would present those twenty to thirty percent matching profiles and give the justification of why those twenty to thirty percent are the right ones to reach out to. And that was really where the majority of time savings comes from too.

David Paffenholz (00:26:21):
Yeah, so the next step was really making the product work. There were some inherent limitations we had because of cost. It cost us a lot of money to serve the profiles. Each user could only access, I think in the beginning, fifty profiles per month, which is a ridiculously low number for a recruiter who probably reviews thousands of profiles in a given year.

Pablo Srugo (00:26:39):
Just because your token costs were so expensive?

David Paffenholz (00:26:41):
Token cost was one, but then we paid different data partners money for every profile that we showed, and we just couldn't afford to show more than we were at the time. And so we fixed that. We got to a point where we could license the data at larger scale and host a lot of that data ourselves instead, which was a big unlock and meant that we could allow users to search as much as they wanted to. And that took us a while. Over the next six months, we were just working on that, trying to get the platform to a usable state. I think revenue at that point was $200k in ARR or so, but it was this very churny $200k ARR. Basically flat, maybe growing a tiny bit here and there, but no meaningful change and that carried all the way through to January 2024, which is when the product started getting good. So for that six-month period, we were just working, making it better, making it better, and it was still just churny.

Pablo Srugo (00:27:31):
And you were funded? How much did you raise at that point? We have tens of thousands of people who have followed the show. Are you one of those people? You want to be a part of the group. You want to be a part of those tens of thousands of followers. So hit the follow button.

David Paffenholz (00:27:45):
We had raised a million at that point. So $500k as part of our YC round and $500k after YC.

Pablo Srugo (00:27:50):
Who was part of the team?

David Paffenholz (00:27:51):
Just Ishan, my co-founder, and me. So just two of us. Yeah, and that was really good because we were very fast on our iteration cycles, and we only focused on what mattered. And so I think we had that forced focus by only being two people.

Pablo Srugo (00:28:03):
Is this something that, I've seen it in many YC founders, that they keep the team very, very small. I'm a huge fan of it, but I'm curious now that you say this, is this something that's reinforced internally by YC, or is it just complete coincidence? Because you had a million, I mean, you could have been, I don't know, six people if you wanted to.

David Paffenholz (00:28:19):
Yeah, so I think YC does say, only hire when it makes sense or when you know what they'll be doing. So they kind of, I guess, softly discourage hiring before the company is sure about it. But yeah, we definitely took that to the extreme. I think part of the reason was we felt we couldn't attract the type of talent that we wanted to hire because we were so early, risky, and unproven. And so I think that was a big influence as well.

Pablo Srugo (00:28:39):
I think, if I remember, Ashby, Ben was on the show, and I think he was similar, he was three people for the longest time, just the co-founders, just building till they got the thing. And to me, it's just a bit of a no-brainer. I mean, when you're in it, you always feel like, oh, if you just had more people, you could move faster, but then you realize, until you know which direction, you're really going to do more stuff, but it's not going to actually get you closer to product market fit in the vast majority of cases. The co-founders think differently, they're invested differently, they're complete self-starters. And it's hard to get that from employees unless you have a direction of where you really need to go.

David Paffenholz (00:29:17):
Yup. Yeah. And I think a lot of that also comes from decision making, where as co-founders, the decision is made together. So both people are automatically bought into it. Versus when there's a team, unless they're also co-founders, the team is not going to be making the decision or have the same influence on the decision. And so it can be a lot harder to get them bought into what the direction should be, especially if that's changing frequently in the early days.

Pablo Srugo (00:29:38):
What happens in those six months where, like you said, in January 2024, things finally start to click? What were some of the key things that made the product really click?

David Paffenholz (00:29:47):
We were still purely focused on search. So our only focus was how good we could get the search and whether this is something that a recruiter would prefer to use over what they've been using day to day. And that was our only North Star. We were just looking at how many searches people were running and whether they were finding the right people. In fact, we got so extreme with it that we had an internal Slack bot that pinged us every time a user ran a search. Then we would manually go and recreate that search. At some point, that became a hundred searches a day that we would go and manually recreate and then assess whether we served the right profiles for that search or what we could be doing better.

Pablo Srugo (00:30:20):
Well, that's what I was going to ask. How do you measure the right profile? So you're doing that qualitatively yourself, literally looking through and being like, yeah, I think we did well here, or I didn't, or I don't, whatever.

David Paffenholz (00:30:30):
And many times it would be obvious what went wrong. We would instantly know the next error point. We added it to our list of things to fix, and then we would go and fix them. Because search is so broad, there's such a long tail of things that can go wrong. The only way of fixing them is methodologically going through every single one of them and fixing it. That takes time, it's also not fun work, but it was necessary. I think we did a good job of executing that step by step. At some point, it crossed a threshold of this is actually really adding value from our workflow. This is a better search than I would get otherwise, at least for some users. We got to that point in early 2024, where people started sticking with the platform and were actually using it as one of their primary ways of finding candidates. We started seeing that reflected in our customer and revenue growth as well. I think the biggest thing that actually changed was not the net volume of new signups. It was the retention of those users. And so we started growing.

Pablo Srugo (00:31:26):
Do you remember the retention? Now I'm going to take you back. What might the retention have been in those first six months, and then how good did it get in the first six months of 2024? Over that twelve month period, what was the change in retention?

David Paffenholz (00:31:40):
Yeah, so our net revenue retention month on month, I think in the beginning, I don't know the exact numbers, but I'd ballpark it at seventy to eighty percent month on month, which is pretty brutal. It means a significant share of users are churning out every month.

Pablo Srugo (00:31:54):
Like you're churning twenty percent of your revenue each month.

David Paffenholz (00:31:56):
That's right. We were roughly adding the same thing back. So overall revenue stayed the same for that seven-month period. Then that started changing in 2024. That net revenue retention number month on month started hitting one hundred percent, and at some point it started crossing that threshold and going significantly above that.

Pablo Srugo (00:32:14):
Because your customers were adding more and more seats. Is that the way you charge, or doing more searches?

David Paffenholz (00:32:19):
Exactly. And so the expansions were always important to us because it showed that it kind of rolled out to a team. But nowadays, with a larger customer base, we take a few different slices of that data and focus on what segments we are winning the most on and where we can still be doing the most work.

Pablo Srugo (00:32:33):
But originally it was kind of like product growth, right? People were just adopting it, signing up, using it.

David Paffenholz (00:32:38):
Yeah. To this day, almost all of our growth is purely inbound from word of mouth, by far our biggest channel. Over half of our signups come from word of mouth. We get over five hundred free user signups a day. Nowadays, we do other things as well. We do paid ads. We just started building an outbound team. So it's diversified a bit, but up to the ten million ARR mark, it was very driven by PLG and then kind of sales-assisted upsell for larger customers.

Pablo Srugo (00:33:04):
Talking to me a bit more about quantitatively just how much better you were. Right. So for example, by the time, let's call it Q1 2024, when things are starting to hit and your net revenue retention is kind of over one hundred percent, people are liking it. When you think about the before and the after, right? So before, they're doing searches on LinkedIn Recruiter or a million different other tools, whatever. Then they're doing searches on Juicebox. What are we talking about? Are you talking about half the time, a tenth of the time, ten percent better searches? How much better were we talking about then?

David Paffenholz (00:33:32):
Yeah, so the initial way we'd measure was, I mean, one, it was very anecdotal, just from what people would tell us about the time saved. And there were two things people told us. One, oh, I found these candidates that I'd never seen before. I've been doing the search for X amount of time, maybe weeks, and I've never found these certain people. That was because our search worked differently than how traditional search worked. And that's very valuable because if you end up hiring one of those people, that adds a ton of value to the business. The second thing is just general time saved. Let's say I used to spend fifteen hours a week searching through profiles. Now I spend five hours doing the same thing. That's also somewhat quantifiable. But there's actually pretty large variance based on who the customer is and how they describe it. Even to this day, it's hard for me to say exactly how many hours we will save for a given recruiter, because so much depends on their workflow, what they were doing previously, and how efficient they were with things before. But yeah, those are the two things that we immediately realized as a good case for ROI.

Pablo Srugo (00:34:28):
It's interesting, right? Because we talk a lot about how you have to be ten times better and all these sorts of things. So much of this stuff, so much of the cliche advice, depends on the go-to-market. In your case, because it's PLG, the user is the buyer, at least that's my understanding, or at least at first, right? Any sort of delta in product quality, they'll feel immediately. It's not like they have to convince somebody higher than them that this is delivering XYZ ROI. Maybe this is a stupid example, but take Slack versus email. How much time do you save or did you save, especially when Slack started? Questionable, but you certainly enjoyed it more and thought it was better. And all of a sudden, you were paying for it. This falls a little bit into that camp where if I'm the recruiter, spending a lot of my time on these tools, if Juicebox works a little better, then that's worth it. That's just worth it to me, and I don't need to over-quantify what that means. Is that about right?

David Paffenholz (00:35:20):
Yeah, I think that's true. I think especially in those early days in 2024, even just a little marginally better was already a big unlock because it meant that this new workflow gave a better output than the existing workflow, and the new workflow had not even been optimized yet. The ten times better always has to be caveated with how important the thing being worked on is. If it's a really important thing, then two times better is a huge outcome for the end user, let alone ten times.

Pablo Srugo (00:35:51):
It's true. This is their core job. I mean, a recruiter literally hunts for talent, and you're going to help them do their job marginally better. There is a point at which it becomes, how can I afford not to be ten percent better than everybody else, or at least as good as everybody else, because they're doing this thing. That makes sense. Walk me through another thing that I really want to understand, which is you mentioned that in some cases you were giving results others weren't. Can you walk me through why that would be or how that would happen? Can you walk me through an example? You mentioned your search works differently. So what's happening, and what might you surface that others would miss?

David Paffenholz (00:36:23):
Yeah, so as an example, let's say we want a pretty basic software engineer who's worked in fintech. The traditional way to set up that search is one of two ways. Either you look for the job title software engineer and then a keyword on the profile fintech. That will pull everyone who has mentioned fintech somewhere on their profile and has the job title software engineer. That, of course, limits you to people who actually have the word fintech on their profile. The second approach is the same method, but also adding a bunch of company names that we know work in fintech. We might research the top one hundred fintech companies and search for anyone who has the word fintech or who's worked at one of those one hundred fintech companies. That second approach already takes a good amount of time researching those one hundred fintech companies. But more importantly, it's not comprehensive. Juicebox takes the software engineer search and actually reviews every single person or every single company to see comprehensively if it matches fintech. That could be purely based on a website describing what the company does, or even a description of someone working at a payroll company but focusing on processing payroll, which probably overlaps closely with what we might be looking for in fintech. Because Juicebox uses a pure LLM for that, we're able to make a suggestion that this person probably matches what you're looking for, even if they never mentioned the word fintech or one of those top one hundred fintech companies.

Pablo Srugo (00:37:45):
Or even if their industry, because you know you have industries as well on LinkedIn, but maybe their industry is not fintech or finance or anything related to that, but it'll still figure it out.

David Paffenholz (00:37:52):
Exactly. And so the reason that's particularly important is because recruiting is so competitive. It's a zero sum game. Only one company is going to hire the right person. And so if you have an edge or a way of finding additional profiles that a different recruiting team does not, that means you're going to be winning those candidates against them because they don't even know they exist.

Pablo Srugo (00:38:09):
You know, one of the things that's really interesting is you're taking, and this is common to LLMs, but the old way of searching is really keyword based. So you either have the keyword somewhere or you don't. Either you have it or maybe the company that you work at has it in their about page. Somewhere you've got to have that FinTech keyword or some way of connecting that this is FinTech, because there is no true semantic understanding of any of these things. There's just what's there on the page. Whereas with an LLM, similar to a human, you can read a profile and think about it a little bit and say, you know what, actually, this is actually FinTech related and here's why. You see LLMs do this type of thing all the time. So you're taking a process that is manual, that has to do with natural language understanding, whatever you want to call it. With LLMs, you can do it at scale, and that gives you this unlock of matching capability that just can't be done if you're limited to the words on the page.

David Paffenholz (00:38:59):
That's exactly right. And it's also exactly what a human is actually pretty good at, but a traditional search system is really bad at, which is looking at a profile and making a reasonable judgment. Does this actually match the search query? And because we can just brute force an LLM to do this at the scale of hundreds of thousands of profiles, we can suddenly use human level judgment against profiles at a scale that was just not possible previously.

Pablo Srugo (00:39:22):
Who is mainly using you? Is it in-house recruiters or like recruiting firms or equal?

David Paffenholz (00:39:27):
In the beginning, because we worked best with, you know, we were just getting started, it was easier for us to enter a small talent team. Often that was startup talent teams, in-house recruiters, or small recruiting agencies. Nowadays, we have over four thousand customers. Sixty percent are in-house, forty percent are agencies. And that ranges from Fortune 500 companies with one hundred-person recruiting teams all the way to startup founders doing their first hire.

Pablo Srugo (00:39:50):
So your ACVs are kind of all over the place?

David Paffenholz (00:39:52):
Yes.

Pablo Srugo (00:39:54):
So tell me about what is the rest of the year 2024, look like in terms of product and in terms of growth?

David Paffenholz (00:40:00):
Yeah, so we kept on growing very consistently twenty to thirty percent every month. We started working with one of the large AI labs in the summer of 2024. I think it's one of the best examples of, you know, they're in a very intense competition for talent and it matters a lot who finds a given profile that another competing company might not. And so we started working with a small team; there were just ten recruiters using us in the beginning, they were kind of testing it out, and they were, you know, looking at every tool in that market, which is also a very competitive process, which we ended up winning. And so that was a really good signal for us that we were, you know, continuing to build in the right direction. And we started winning more customers around that time as well and started working with teams too. Teams opened up a kind of new problem set for us. In the beginning, we were just solving for the pure search; now we have to solve for the search in the context of many other people at the same company also running the same search. And so at what point should one serve the same results versus when should results be different? How does one hide views or activities that other people on the team have done, and more? And so we started solving for that problem landscape as well, while also solving for the deeper workflow. So everything from managing those contacts to engaging them, reaching out, and more. And we made our first hire on the team. Our founding engineer, Minju, joined us in the summer of 2024.

Pablo Srugo (00:41:13):
What revenue were you at when you made the first hire?

David Paffenholz (00:41:16):
We were close to a million. I think we were at like $700K or so, but growing pretty quickly. So, like, I think maybe within two months or so of him joining, we hit the million.

Pablo Srugo (00:41:25):
And this was mid-24?

David Paffenholz (00:41:27):
That's right.

Pablo Srugo (00:41:29):
So I'll call it a year and a bit ago. How many people now on the team?

David Paffenholz (00:41:31):
Twenty-four.

Pablo Srugo (00:41:32):
So you kept the team relatively lean. I mean, again, you've raised more millions than you have employees.

David Paffenholz (00:41:38):
I don't know if that's KPI, but yeah, and our revenue per employee is still very high too, which I think is a good metric on team skilling.

Pablo Srugo (00:41:45):
Let's go deeper on go-to-market. I mean, at this point, you've got a product, you know, it's generally working, you're expanding the product features as you normally would. Inbound is a funny thing. It is great in one sense because you don't have to do anything, but it also has a potential downside—not always, but a potential downside—which is it's just not as controllable, right? Like outbound, if you told me, you know, we went outbound, I'd be like, okay, so you probably what, you just hire more SDRs, you find a way to get more leads, you get more AEs, you just scale the outbound team, you scale the top line, boom. With inbound, the worry is, you know, whatever flow you're getting now, you're going to have some word of mouth, it's going to add a little bit, but is it going to double? Is it going to ten times? Like, how do you just pull on that dial? So tell me maybe a little bit of how you structured go-to-market over the last twelve months.

David Paffenholz (00:42:27):
Yeah, so all throughout 2024, it was founder-led sales. So I was doing sales, winning the year at like 1.5 or so million in ARR. At that point, I knew our customer really well, because I'd just done so many sales processes. If you booked a demo with us, you'd be booking a demo with me.

Pablo Srugo (00:42:42):
Yeah, what was maybe, so they would go on inbound, book a demo, and then you talk to them and get the close. That was kind of the funnel.

David Paffenholz (00:42:48):
Exactly. And it would be, we'd run like a one-week trial process for larger teams. And we did a ton of variations of that, from a four-week trial to a three-day trial, and then we settled on the one-week period.

Pablo Srugo (00:42:58):
And $50 a month? That was the price point?

David Paffenholz (00:43:00):
At that point, we had increased it a bit. I think we were around $100 a month or so.

Pablo Srugo (00:43:05):
Was there a full self-serve, or did you have to do a demo?

David Paffenholz (00:43:07):
Yes. In fact, the far majority of revenue came through full self-serve.

Pablo Srugo (00:43:11):
Okay.

David Paffenholz (00:43:11):
But then, of new revenue, over time, the kind of expansions and some of the larger customers ended up becoming an increasing share of our customer base. Today, it's the majority of our revenue.

Pablo Srugo (00:43:22):
So to be clear, when you're saying you're doing demos, some people were booking demos, you were doing those, but also some people were just signing up inbound and using the product.

David Paffenholz (00:43:28):
Exactly right. In fact, many times, people would book a demo after they already purchased their first monthly plan. So they'd be like, you know, I bought this, I wanted to get started right away. I think we have a pretty fast time to value. So from when you test out the product to when you see the value in it, it's pretty quick. Our trial is also pretty limited. And so at that point, it's kind of a purchase decision or the book-a-demo decision.

Pablo Srugo (00:43:47):
How big of a deal has time to value been? This is a KPI I keep hearing about. And I just, more and more, I'm like, this is one of the biggest things you can focus on in the early stage. I'm curious what your impression has been.

David Paffenholz (00:43:56):
Yeah, I think time to value is like a requirement for doing a successful self-serve motion. For us, what that means in the product is that today, when you sign up to the product, we put you right into a search. And so the first thing you see is like the prompting interface to put together your first search. And then it's literally just one click from there, like clicking run search to start seeing profile results. And oftentimes, that's already the value for our customers. They quickly got to people they might want to hire, and it can go as fast as like five seconds. And that is really important for us because the attention spans are short post sign-up. Like, what is the value we're actually delivering? And oftentimes, software has like this initial empty state where there's nothing in there. There's no value that exists because you just created the software. And so we focus a lot on just making that time to seeing your first search results really quick.

Pablo Srugo bn (00:44:43):
How do you make that? Is that search personalized at all, or is it just like a template search that you have loaded out?

David Paffenholz (00:44:47):
It's an open prompt field, but we have some pre-filled searches that you can just click, or you can write your own. Usually, when people sign up, they sign up because they have a search in mind. Like, they know they want to hire for someone. And our example searches are also intentionally short so that the user feels they're also fine with just writing an initial short prompt to get started rather than something super comprehensive.

Pablo Srugo (00:45:12):
Perfect. I took it completely off tangent because you were talking about go-to-market. I took you down the time to value road.

David Paffenholz (00:45:17):
No, the two go hand in hand. So like, the time to value has remained quick as the customer gets larger, and it's a larger organization. When selling to time to value alone, it's not enough to get a conversion, and there's a lot more depth behind it too. And so I think that was kind of a learning journey we've gone on this year in 2025. Our first account executive joined right at the start of the year in January, and today we have six account executives on the team. They do exactly what I did, and they do it better, which is getting to know the customer, seeing what their needs are, and then running a one-week trial process with them. We're still very trial-heavy. I think there's a lot of software that says no trials, or it's just a purchase process. That's, I think, where the time to value still comes in. We're able to show strong results in that one-week trial period, and it's kind of still a core part of our selling motion. And now the difference is that because we work with larger teams and there are more users, the contract values are accordingly higher.

Pablo Srugo (00:46:09):
And you've grown what, like 3x this year?

David Paffenholz (00:46:11):
This year closer to 10x.

Pablo Srugo (00:46:13):
Oh, wow.

David Paffenholz (00:46:13):
Okay.

Pablo Srugo (00:46:13):
So you've crossed 10 million top line.

David Paffenholz (00:46:15):
Oh yeah. We announced, so $10 million, we announced as part of our Series A announcement.

Pablo Srugo (00:46:19):
That's an insane year. So tell me a bit about the top of funnel because, I mean, inbound, like, you can add AEs and all that stuff. Obviously, it's critical, but that's post somebody signing up or post somebody booking a demo. How do you just 10x your top of funnel on something like inbound like this?

David Paffenholz (00:46:33):
Yeah, so word of mouth has continued to grow. I don't know if it's actually grown 10x, but it's continued to grow pretty aggressively. And then at some point, we just started layering in other channels. And so Vicky on our marketing team joined us in February, and she's been just experimenting and driving so many different things. Everything from pretty large-scale paid ads to, we did our first billboard campaign recently, here in San Francisco. A lot of billboards on Juicebox, we host events, we go to the different recruiting conferences. And so we've been testing all these different channels just to layer on and see what works, and then double down on the ones that do work for us. And some of the ones that we found work are: one, organic content. And so we post a good amount of content as a company, be that like posts that I post on my LinkedIn, but also other team members and on the company page. Two, we get very involved in recruiter communities. And so we might sponsor them, we might invite them to events, we might co-host events with them. And that's really just to amplify the word of mouth. And then third, paid ads. And paid ads serve two purposes: one, kind of acquiring new customers, but then two, engaging people who tried it in our free trial but never converted yet. And so retargeting is also a big component of that for us.

Pablo Srugo (00:47:40):
I think if I asked like ten founders who they would want to lead their Series A, Sequoia would have to be on the top of just about everybody's list. How did the Series A with Sequoia happen? Can you share that story?

David Paffenholz (00:47:49):
Yeah, it happened fairly quickly. So David Kahn, who is now on our board, reached out to us in the summer, and we got an intro through an angel investor in Juicebox and a power user of Juicebox who connected us. And then we knew we kind of wanted to raise an A, but we weren't sure on timing yet. And so we kind of took the conversation just with like getting to know each other and…

Pablo Srugo (00:48:09):
It wasn't part of like a process. It was just like.

David Paffenholz (00:48:11):
Yeah, there was no process. And so then, after a couple of conversations within a one-week period, it was clear that they were leaning in. And so we decided, okay, now, if we want to continue these conversations, we should make it a process because we don't want to kind of go down the road with just one firm. And then, you know, who knows what the outcome is. And so we had spent that initial week with just Sequoia, and then in the second week, we kind of started creating a deck and making an actual Series A process. And that's where we then spoke with a targeted number of other firms that we had built relationships with in the past and that we'd also be excited to work with. But we kind of knew that if Sequoia would work out, that's where we wanted to go. And then fortunately, we got the term sheet the following Monday and then kind of agreed on those terms shortly thereafter.

Pablo Srugo (00:48:56):
Tell me, when did you raise your Seed round?

David Paffenholz (00:48:58):
We raised the seed in the fall of 2024. So, like, October, and at the time it was just Ishan, me, and our founding engineer, so the three of us.

Pablo Srugo (00:49:05):
How big was the Seed?

David Paffenholz (00:49:07):
It was $5 million.

Pablo Srugo (00:49:08):
How much of it did you really use between then and raising your A?

David Paffenholz (00:49:11):
Almost nothing. I think like, seven hundred fifty thousand, five hundred thousand or so. So we did not burn much during the time. And a lot of that was just due to the revenue growth and the fact that we had, no, we have pretty high margins, or we have no huge cost structure. And so I think that allowed us to be fairly efficient as a business.

Pablo Srugo (00:49:28):
There's many different philosophies on raising money. You decided to raise thirty million dollars, even though you hadn't spent the five. What's your thinking for why that makes sense?

David Paffenholz (00:49:36):
Yeah, I think part of it is, each firm will have ownership that they want to get to. And so part of it is a trade-off: if you want a sufficiently high valuation, that also means the round has to be of a certain size. And so I think that's one element of it. And then the second is, you know, when the company is in a position of strength, it can make sense to kind of build that capital. And then in an ideal case, you never need the capital. But a startup journey is never, you know, just up and to the right. And so it can also be good to have a bit of a buffer in case things are at any point not up and to the right.

Pablo Srugo (00:50:05):
Perfect. Well, listen, let me stop it there, and I'll ask the three questions we always end on. The first one is, when was the moment when you felt like you'd found true product-market fit?

David Paffenholz (00:50:12):
I think when we signed the AI lab as a customer in mid-2024, and they expanded shortly thereafter. Maybe the expansion was actually when we knew we had product-market fit because we knew we could sell to one of the hardest-to-please customers. And then, they actually liked the product and expanded with us shortly thereafter.

Pablo Srugo (00:50:32):
Was there ever a time on the flip side where you thought things might just fail?

David Paffenholz (00:50:35):
Yes. After we launched, we had that viral People GPT video launch in 2023. We had all these users, but then three weeks later, they were all unhappy or turning and wanted a refund, etc. And that was really rough because we had just come off this high of, you know, everyone is interested in the product, or we had our message-market fit, but then we felt like we couldn't deliver on it. I think that was kind of a rough time.

Pablo Srugo (00:51:00):
And then what would be your number one piece of advice for an early stage founder that's trying to find product market fit today?

David Paffenholz (00:51:07):
Keeping the team super lean and then being honest with yourself. Nowadays, one of our core values is intellectual honesty. And I think early on, when we kind of wanted things to work one way, we would lie to ourselves a bit. And I think we always ended up regretting that or making a decision to reverse that later on. And so, trying to be honest with oneself and trusting your gut instinct of, you know, this doesn't feel quite right.

Pablo Srugo (00:51:30):
Awesome. Well, David, thanks so much for jumping on the show, man. It's been great having you.

David Paffenholz  (00:51:33):
Thanks for having me.

Pablo Srugo (00:51:34):
Wow, what an episode. You're probably in awe. You're in absolute shock. You're like, that helped me so much. So guess what? Now it's your turn to help someone else. Share the episode in the WhatsApp group you have with founders. Share it on that Slack channel. Send it to your founder friends and help them out. Trust me, they will love you for it.