This Ugly Website Does $106M/Year (866,000 Pay For It)

Vadim Yasinovsky, co-founder of pdfFiller and CEO of airSlate (parent of pdfFiller, signNow, DocHub, US Legal Forms, airSlate Flow), breaks down how a plain-looking PDF editor page became a $106M/year product with 866,000 paying customers — inside a company doing $200M in revenue across 1.1 million customers at an average of $180/year. He raised just $150M total (last round: $52M at a $1.25B valuation in June 2022) and now says the company is optimistically worth $800M — despite doubling revenue since. He bought signNow at $3M and grew it to $35M. He built an SEO machine on 4.8 billion keywords and 35 million pages, and now says that playbook is dead within 9–12 months as ChatGPT and AI search take over. In this episode: Why 650 employees are split into siloed 120-person product teams with no shared tech Why airSlate didn't become Harvey despite 85,000 lawyer-built legal forms as a training set Why Vadim says the "ChatGPT SEO" gurus are "b.s." and his influence on search drops from 80% to 30% His unannounced AI product (bizdraft.com) and why he won't monetize it The path from $200M to $400M: 20–30% from existing products, the rest from vertical AI His M&A filter: product-market fit, domain expertise, $5–10M+ revenue minimum His advice to founders: raise as late as you can, cash is king, preserve it now Topics: SaaS, SEO, programmatic SEO, AI search, ChatGPT, document automation, e-signature, PDF editor, legal forms, SMB software, bootstrapping, vertical SaaS, M&A, capital efficiency.

• 0:00Speaker 00

  Our customer, and there are 31 million of them in the US, is SMBs. It's the guy who fixes your plumbing, who bakes your bread, who sells you books in a small bookstore. They don't need complicated agreements, which will withstand scrutiny of multiple courts and multiple states and multiple regulations. And I want to answer 10 questions and get this shit done.

• 0:22Speaker 01

  I think people would be shocked if they landed on this page and they knew that this page right here does $106 million of revenue.

• 0:29Speaker 00

  DocHub now is doing 20, 22 million, something like that. PDF Filler is doing about 106 and Cyan now is around 35.

• 0:38Speaker 01

  All these gurus right now saying they've figured it out, but I don't know that it's actually

• 0:42Speaker 00

  No, no, no, it's bullshit. I'm sorry. I've tried this stuff.

• 0:47Speaker 01

  Hey folks, my guest today is Vadim and at his request, I'm not going to try and pronounce his last name, but he started his first company in 1982 and after several startups co-founded PDF Filler, which grew into Air Slate. He's a longtime software industry veteran and returned to the Air Slate CEO role in July of 2025, now leading the company's next generation of AI products. Vadim, you ready to take us to the top?

• 1:11Speaker 00

  Yeah, well, you know, next generation is actually a very good beginning because we've been doing on and off AI for like 40 years. I'm not exactly a new guy around the block. So our approach is AI best when it's invisible. AI best when it actually produces real value. So we're not putting labels here and there saying AI this and AI that. We're just trying to do best job for our customers.

• 1:38Speaker 01

  You've acquired many of my friends' companies, right? Chris at DocHub, you guys did that deal. Obviously you were the first, I think, anchor, but you've let all the brands you acquired still keep their original sort of name and structure. How does that impact your AI strategy going forward?

• 1:53Speaker 00

  It's actually really interesting because the process of the different stage on their development and their product life. PDF Filler is primarily prosumer and we're getting a lot of traffic and the conversion is very good. DocHub really needs some work in marketing and Cyan now is a brand sort of recognition game. The interesting part is that while they're different, we're using same technology for the customer acquisition, but fundamentally the idea was that we have small amount of common services that include marketing, you know, things like DevOps, payments and things like that. But the rest of them are siloed. They have their own GMs, they have their own development team, they have their own

• 2:37Speaker 00

  roadmaps and they sort of go in their independent way. And so far, so good. I don't think AI will actually change anything in that sense. We do have something called AI Labs, which develops common AI solutions, which we shift into the products, but there's no mandate to adapt to any technology developed in house. Cyan, now PDF Filler and DocHub are three entirely different products, different tech. There are some common components which they share, but very few. We sort of made it intentional because we honestly believed, especially in our days, tech is not the mover. It's the brand, it's the, you know, it's the top of the funnel. That's, that's the things which differentiate.

• 3:20Speaker 00

  In reality, we're not going to even try to adapt it to the same tech because economy of scale doesn't work anymore with AI.

• 3:27Speaker 01

  How many people are full time today across the whole company?

• 3:30Speaker 00

  There's about 650 people in the company right now. So products have about 120 per each. The rest of it is GNA, marketing and common services. DocHub now is doing 20, 22 million, something like that. PDF Filler is doing about 106 and Cyan now is around 35.

• 3:50Speaker 01

  I think people would be shocked if they landed on this page and they knew that this page right here does $106 million of revenue.

• 3:57Speaker 00

  Hey, you know, we have close to 860,000 paying customers. So yes, right now it's $100 million product and we hoping that it will grow to maybe 200 million within about two to three years.

• 4:12Speaker 01

  What's the rest of the breakdown? So DocHub is 22 million, PDF Filler is 106, Cyan now is 35 million. What about the other three?

• 4:19Speaker 00

  So Out of Flow is sort of a very, it's offshoot for us. We've been at that thing for quite a while now. It's not a big product. It's about 5 million revenue, something like that. US Legal is about, I think, 6 million, something like that. It contributes a lot of traffic and a lot of value to PDF Filler. And it's actually, there's like about 85,000 forms, which are hand-built by lawyers. And it's a humongous asset because it allows us to use this data to build new products. We actually have one more product, which you do not have on the list because we haven't really publicly announced.

• 5:00Speaker 00

  The URL is bizdraft.com, B-I-Z-draft.com. And it's our first AI product. We build it specifically for our users. It creates agreements, legal agreements. The interface is unique because we actually combine editing and AI. And we've seen some really interesting stuff about the traction so far. I'm very pleased. Most likely, we're not going to actually monetize that. It's much easier for us to use this as a traffic generation engine.

• 5:34Speaker 01

  I want to ask you like a deeper question here, right? So let's just look at US Legal for a second. This site structure is actually very interesting because they have these thousands of templated forms, which pull in a lot of traffic. And I'm not even sure that Ahrefs is capturing it all, but I'm leading to a question. My question is this, Harvey and all these legal AI tools, they have to have a large training set to sort of win. You guys should have won that space because you had a very large training set. Why weren't you guys Harvey?

• 6:01Speaker 00

  You have to remember, our customer, and there are 31 million of them in the US, is SMBs. It's the guy who, you know, who fixes your plumbing, who bakes your bread, who sells you books in a small bookstore. And these guys don't need this monstrosity and huge building sets. We have enough information to build a solution, which is, I wouldn't say good enough. I would say great enough for them. Okay, they don't need complicated agreements, which will withstand scrutiny of multiple courts and multiple states and multiple regulations. And I want to answer 10 questions and get this shit done. Okay, and this focus on the customer is extremely important.

• 6:42Speaker 01

  To remind me, how much capital total have you guys raised today?

• 6:45Speaker 00

  I think about 150, if you count, you know, all the loans and stuff like that. But I mean, we got money from GC Capital, of Morgan Stanley, G Square, and some minor,

• 6:58Speaker 01

  UiPath Ventures.

• 6:59Speaker 00

  Yeah, UiPath Ventures. Yeah, it was a minor investment, not that big of a deal.

• 7:04Speaker 01

  That last price round though was June 2022, 51, 52 million raised at a 1.2 million, 1.25 billion valuation, right?

• 7:11Speaker 00

  That's right, that's right. Your numbers are better than mine.

• 7:16Speaker 01

  I have good memory, but how do you, are you, I mean, what are you seeing in the market today? Do you think if someone offered you, I mean, are you in the money today? You think you're valued north of 1.3 billion today?

• 7:25Speaker 00

  No, I don't think so. I mean, let's be honest, market really went down. You know, I would place the value of the company right now. I mean, optimistically, maybe 800. And you have to understand, it was $100 million revenue when we got the last evaluation. Now it's 200 million revenue. So the reality is a very substantial. I think my market is panicking. I think my VC friends really don't understand what's going to happen. I think I have a very clear vision of where we're going and how it's going to play out. I do not believe for a second that my customers are going to build their own software. That's absolute baloney. And only people who never build products can actually think that that's what's going to happen.

• 8:06Speaker 01

  Can you fill in those gaps really quick? So the metrics I have from direct interviews with you and Borya was 2022, 100 million and 2026, 200 million. But fill in the gap between 2016 and 2022, like when did you break 50 million?

• 8:20Speaker 00

  2015, I would say we got the money from GC. That was the first investment of capital. We bought Cyan now, we rebuilt Cyan now, and we, we got to about 50 million.

• 8:31Speaker 01

  Did you break 10 million before the money from GC in 2015?

• 8:34Speaker 00

  Oh yes, absolutely. We raised money when we were close to 18 and a half million. That's, that's, that's, that's the point.

• 8:42Speaker 01

  So, so you hit 18 million of revenue in what, 2014?

• 8:46Speaker 00

  Yes, something like that, maybe 2014.

• 8:49Speaker 01

  And when you bought Cyan now, that basically doubled the business? Were they doing about the same revenue as you?

• 8:54Speaker 00

  No, it didn't double the business, far from it. We bought it, it was doing about $3 million a year. Now it's about 35. So it didn't really, I mean, we had to rebuild the entire product and we had to, you know, weaponize the marketing stuff. So it took a while. No, PDF Filler continued to doing great and that's where we got most of the money. I mean, the way company was built, PDF Filler was a cash machine and we just take the cash and invested in other products.

• 9:21Speaker 01

  When you guys sit down and have a coffee and look back at the golden years, printing money, no BOR, no VCs, do you ever go?

• 9:26Speaker 00

  Oh God, please, you make me feel so nostalgic.

• 9:32Speaker 01

  Would you change anything, though? I mean, teach the founders listening right now. Would you change anything?

• 9:37Speaker 00

  Try to get money as late as you possibly can. Cash is the king. Make money. Don't pay yourself a lot of money. Don't spend money. I mean, and from day one, think about exit.

• 9:53Speaker 00

  That that is something which I learned. I don't like to exit, to be honest. It's like a baby, right? I mean, you've raised this thing. You want to become a real company. You want to build real value. You want to build cool products. But you have to think about exit all the time. And guys, if you can avoid it, don't raise money. I know that it's sort of, you know, a very modern tendency, which is because we need to scale. It's it's funny. I was teaching this seminar on software 101, how to build software company. And I asked one question to the audience. I said, what is the essential part of having a company? And, you know, we will say great business plan, funding, great talent. And they're going on and on and on and on.

• 10:37Speaker 00

  And this little girl in the front row, who must have been 16, said, customer. Said, finally, somebody figured it out. You know, that's what it's all about. It's about customer.

• 10:48Speaker 01

  I think you said 800,000 paying customers today?

• 10:51Speaker 00

  No, actually, altogether, we have about 1.1 million. But PDF filler has 860,000 paid customers.

• 10:59Speaker 01

  So 1.1 million paying customers into

• 11:02Speaker 00

  Altogether, all the products.

• 11:04Speaker 01

  Yep. Into a 200 into a 200 million ARR base means each customer is paying you on average $180 per year. I mean, that is the definition of SMB.

• 11:13Speaker 00

  Right.

• 11:14Speaker 01

  Yeah.

• 11:14Speaker 00

  Exactly.

• 11:15Speaker 01

  So if someone else is running that same playbook, you know, and we dig in, I mean, this is where all your this is where all your top of funnel is coming from. What did you guys do with PDF filler to build such a massive organic content machine?

• 11:26Speaker 00

  We have under management, as far as I remember, 4.8 billion keywords. And we had 35 million pages. This stuff has changed significantly, so it worked really well for us for quite a while. But what's happening right now is very different. AI is getting smarter. Search engines are getting smarter. Our ability to create perception is getting is weakening. Let's put it this way. AI requires a very well structured information. We throw a lot of stuff out there and it's not exactly clean. It's pretty messy.

• 12:08Speaker 00

  So now we're going into a process of reducing what we had before and making it more productive.

• 12:14Speaker 01

  When I came up there to Boston and you, me and Boris, some other folks, you guys got together, we had a brainstorm. I mean, I remember one of the topics of conversation was how to think about contextual relevance. So instead of programmatic SEO like this kind of keyword here, it was how to build contextual relevance around keyword clusters. Is that still your thinking today? And how are you testing this?

• 12:33Speaker 00

  Absolutely. Yes, yes. You know, we're talking about clusterization. We're talking about semantic core. We're talking about discovering new keywords and so on, so forth. Now, you do have to understand that you take this huge, you know, 4.8 billion. Very few of them are actually productive. So figuring out which clusters are more productive and trying to cover as much as we possibly can. That's definitely part of the journey. But I got to tell you, Nathan, this stuff is about to change and it's going to change, not overnight, but it's going to change pretty drastically.

• 13:07Speaker 00

  Our ability to influence and bring people over is going to be probably significantly diminished. If before I would call it 80%, now it's going to be 30%. And it doesn't matter how many landing pages we have. It's the quality versus quantity now. That's the name of the game. It's real sources like Wikipedia, Reddit, things like that, which will make a huge difference. I'm not telling you anything you don't know already. I'm just saying that this massive push towards enormous amount of keywords, I think, has come to the end or will come to the end within the next, I don't know, nine months to 12 months. This stuff is happening.

• 13:48Speaker 01

  And let's use a real example. Okay, let's say I am a pool contractor in Austin, Texas, and I need, you know, and I pay PDF filler because I found a long time ago some template doc about how to get a proposal out quickly to a pool owner in Austin, Texas. Moving forward now, if I'm that same, a new contractor, pool contractor in Austin, Texas, and I go to ChatGPT and I say, free two-page template for me to get my first, you know, proposal out in Austin, Texas. Can you actively influence? Have you found a way to actively influence the results ChatGPT is showing? There's always all these gurus right now saying they've figured it out, but I don't know that it's actually figure out a goal.

• 14:26Speaker 00

  No, no, it's bullshit. I'm sorry, I've tried this stuff.

• 14:32Speaker 00

  Again, think about it. ChatGPT today has IQ of about 110. Once it has IQ of 400, whether you and I have 80 or 140, it doesn't make any difference. It's getting much smarter and it's very unpredictable. So the answer is adaptability. You know, there's no way, we're not going to have the same wonderful scheme of SEO experts telling you what to do. Forget it. One of the reasons why we were so successful, I, you know, I'd be honest, our competition was not sophisticated. So we use high level math. Our bidding on PPC was superior. Now everybody has a clock. We are on an equal basis.

• 15:15Speaker 00

  So you have to be, I think it's going to be a lot more about the product and about reputation in the future.

• 15:23Speaker 01

  What is the growth plan moving forward? How do you go from 200 million to 400 million of revenue?

• 15:27Speaker 00

  Double back on our existing products. We believe there's a, you know, 20 to 30% growth in there. But majority of the plan is probably going to go to new AI products. And by the way, the example you brought up is really interesting, the guy at the pool. Think about this guy not worrying about how to find the document, what to stay with compliance, you know, how to make sure that the pool is going to be clean to the point of the inspection, permits and things like that. That's a huge area where these poor people spend six to eight hours a week. Instead of getting a couple of slices of pizza and a nice cold beer and watching a football, they're sitting there with their fingers on the keyboard, trying to figure out how to fill up the damn document.

• 16:12Speaker 00

  And there's a bureaucrat somewhere who needs to approve it. So this is, we believe it's a humongous market. We'll be able to tackle those problems and we'll be able to provide enough of a value to a lot of people and remain horizontal while verticalizing the product for a lot of vertical sort of submarkets.

• 16:31Speaker 01

  My argument is you really have to go vertical to keep your net dollar retention high, but you want to stay horizontal.

• 16:36Speaker 00

  Now you're talking. So how the hell do you do that? And the answer is your product has to morph itself. It has to speak the language of your customer. So if it's a small medical office, you know, it needs to understand this stuff. If it's a, I don't know, pool guy, it's a different language. It makes all the difference in the world. And I think that AI actually allows you to morph your software to address a particular user segment. And the great thing is that we don't have to rush. We can wait for the feedback loop to kick in. We are not a startup in that sense. All the startups are going to be stars to that before they actually get the real feedback. There's uncertainty in this times. It's coming big time. Guys, preserve cash.

• 17:21Speaker 00

  Do yourself a favor. There's nothing else like that in the world.

• 17:25Speaker 01

  Are there legacy, like we look at what Bending Spoons has done, right? You guys know a thing or two about buying companies.

• 17:30Speaker 00

  Yes, yes.

• 17:31Speaker 01

  So if you were going to run their playbook and you were thinking about acquisition targets today, who would you think about going and acquiring?

• 17:37Speaker 00

  You know, show me success on product market fit and domain expertise and we can talk. And yeah, it has to be more than, I don't know, 5 million or even 10 million, because it needs to make a significant contribution to the revenue of the company. You know, so 1 million products are not interesting.

• 17:55Speaker 01

  You said you hate managing like a board. I think you really like being your own boss, but I mean, you have a real board today. I mean, there's a lot of money in the business. Like, how do you manage that?

• 18:05Speaker 00

  Low expectations, under promise, over deliver, consistency. You know, when I took over, which was almost a year now, in the beginning, they were very, very frantic and nervous. And I just made it my business of talking to them all the time and telling them why I'm doing it, what I'm doing it. And eventually they start trusting me. And now, you know, they bug me on on the board meetings and it's like, it's been great.

• 18:30Speaker 01

  Because even though you have a board, you're still, I would consider you guys very capital efficient. If you're at 200 million bucks of ARR, you've raised under 200 million of equity to date, right?

• 18:39Speaker 00

  Yes, yes.

• 18:40Speaker 01

  Very, very capital efficient. There's people listening right now that have, you know, raised 20 bucks for every dollar of ARR. That's not you.

• 18:46Speaker 00

  That's not us. And at the end, you have to win big. And remember, VCs are in this game for different reasons, right? They want to win big. They can't afford, I mean, if one company out of five wins big, they're cool. And the rest of them, they don't care about. While I respect my board, I like those guys very much. I do have to understand that their objective and my objective are slightly different. And I have to adjust for that all the time.

• 19:13Speaker 01

  On that note, if people want to follow your story after meeting you here on this interview, where can they find you online?

• 19:19Speaker 00

  You can't find me online. I'm not anywhere. I'm building new AI products. So I mean, you can, you can, you can send me an email if you want, with him at earthstate.com. I don't check Facebook. I'm sorry, I'm like off the grid.

• 19:32Speaker 01

  You don't do a lot of interviews. Why'd you agree to do this one?

• 19:35Speaker 00

  So, I mean, I saw you laugh. I could make you laugh. I think we have same sort of foundation basis. This is not an interview to promote the company or anything else. I mean, we can talk about, you know, we can shoot breeze about whatever, whatever the hell you want. Doesn't matter. So that's why I'm here. I mean, I don't do interviews. You're right. I mean, this is like exceptional. So, and you've done a great job, man. A lot of respect.

• 20:01Speaker 01

  Well, I've got a ton of respect for what you guys have built. So I appreciate you making time for us. But Dean, thank you for taking us to the top.

• 20:06Speaker 00

  Thank you, sir. Really appreciate it.

• 20:09Speaker 01

  If you enjoyed that interview, you should click here to subscribe to the channel so you don't miss new interviews every Tuesday. If you've got some free time and want to listen to the next interview right now, click right here. Lastly, subscribe wherever you get your podcasts. We're on Spotify, Apple Podcasts, and here on YouTube.