He got rejected from YC—then grew to $1.8B in under 2 years. | Max Junestrand, Founder of Legora

Max Junestrand (00:00:00) :
We've gone from 0 to $1.8 billion valuation in less than two years, and I said, the only way we succeed is if we serve all these clients well. And if we get from, 0 to $20 thousand MRR or a million ARR. And I went back, I sort of locked myself in for five weeks. I was running around Stockholm with my little briefcase. A lot of travel to Finland, to Denmark, to Norway. We ended the YC batch with almost a million in ARR. So I went over to the computer. I'd tabbed out from the PowerPoint. I go to Google, I log in on Legora. I'm shaking, you know, I've never done a demo for two hundred people before. I show a couple of use cases and everything works perfectly. So I go to our webpage and I'm like, and here's the book a demo. And then my phone starts to vibrate because I get all this demo bookings. So I actually did one hundred fifty demos off the back of that presentation. I had just showered and I was trying to talk not too loudly, because everybody else was sleeping in the room. This is the biggest deal ever, it was $45k and I was, this is amazing and the team was wild. And we celebrated, then they kept coming in. Another $30k deal, another $20k deal, another $40k deal.

Previous Guests (00:01:16) :
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 (00:01:28) :
Do you think the Product Market Fit show, has product market fit? Because if you do, then there's something you just have to do. You have to take out 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. Max, welcome to the show.

Max Junestrand (00:01:44) :
Thanks, Paulo. It's great to be here.

Pablo Srugo (00:01:46) :
Dude, so I mean, I'm looking at Crunchbase here and if I have the numbers right. You start this maybe two years ago, like early 2023, maybe two and a half years ago. You get into YC in April of last year, and then things just go ballistic.

Max Junestrand (00:01:59) :
They go pretty bananas.

Pablo Srugo (00:02:00) :
They go crazy, man. A month after YC, $10 million from Benchmark, then a Series A two months later, $25 million. May of this year, $80 million Series B from General Catalyst, and then last month, $150 million from Bessemer. That is some of the most insane fundraising history I've ever seen.

Max Junestrand (00:02:16) :
It's been a really hectic ride. Somebody wrote on LinkedIn that we were maybe the fastest growing Y Combinator unicorn in history. I'm not sure if that's correct or not. You know, I think what happened was following YC, we had so much traction and we can talk about those early days. But after the Benchmark round, we got immediately preempted with the Series A, and then we really sort of went into GA like October 2024. So until then we had been pretty relaxed about the product. We were sort of pretty stealth, and then we said, hey, now we're ready to onboard a thousand lawyers every day if we have to.

Pablo Srugo (00:02:51) :
Wow.

Max Junestrand (00:02:52) :
And so basically since October 2024, that's when we really started to scale. And this year has been nuts. We were forty people at the beginning of the year and now we're almost two hundred fifty.

Pablo Srugo (00:03:03) :
The ability to scale incredibly quickly in this post Gen AI world, at least just because of the demand. I don't know if we've seen that before. I mean, there were insane growth stories before, of course, but a lot of those were funded by dollars, right? Crazy ad spend, you know, selling a dollar for ninety cents and in some cases it worked. Like, Uber or whatever, but these days it's actually selling profitable products. At least on a gross margin basis and yet demand is just incredible.

Max Junestrand (00:03:28) :
It's also different to scale an enterprise. I mean, Wyss did it, but the Uber model. I mean, it's a very different type of scaling and we do enterprise contracts one to three years. We work with the two hundred biggest law firms in the US and the biggest corporate departments of the world. That's a very different sales motion. It's a very different expectation of your product. But I think the difference is, these products have not been possible to build before, right? What we're coming in is not a displacement of existing software. Well, sometimes we displace competitors, but we're coming in with something new that they haven't done before.

Pablo Srugo (00:04:04) :
So I mean, we're going to dive into all this because in your case, the history is not that long and so that's fortunate. Because we can actually get into all of it and one of the things I want to really understand. Because to me, it's fascinating, is never mind the demand. It's one thing to tap into crazy demand. It's another thing to even just fulfill it. I mean, just to fulfill it at that rate must be insane inside. So we'll get into all that. Let's maybe start at the beginning. Where does the idea for Legora come from? What were you doing at the time and can you walk us through that origin story?

Max Junestrand (00:04:33) :
So this is going to sound like the beginning of a great joke, and I'm actually not sure if I've told this story publicly before. But the founding seed of Legora was started in 2021, and it was a lawyer, a physicist, a psychologist, and an engineer.

Pablo Srugo (00:04:52) :
That is the beginning of a joke, yeah and they go to a bar.

Max Junestrand (00:04:55) :
And I was not, you know, part of that crew. So it's four guys in 2021, and it was August when they pulled the thing together. They basically oriented themselves around this idea that the lawyer had so many boring tasks when he was doing an internship in a courthouse, summarizing court cases, finding similar case law and things like this. They took the early BERT models, which was sort of pre GPT, GPT1, GPT2, and threw themselves at the problem. Trying to build cool stuff on top of it and they kind of failed. It was really hard, especially in non English languages. In Swedish, the models were just extremely unsophisticated. Then, in the post ChatGPT era, when the APIs became available on GPT 3.5. We revisited the problem, and they built a really cool demo basically. They showed it to me, and it was around how to make sense of your stock option agreements as an employee. Because it's a lot of legalese and hard to understand. The lawyer and the physicist decided that this is not for us and decided to go back to the courthouse, and the physicist went and did quantum computing in Amsterdam. Then I joined, and the reason why I joined was because August and Sigge wanted somebody who had gone to business school. Because you needed somebody who had been to business school to go to the business accelerator, and I'm not really a business guy, I went to engineering and then managed to add on business as an extra thing. So I just said, well, put my name on it if you want to, and then make an application. That sounds fine, and I got a little bit involved and said, why don't we apply to Y Combinator? That sounded a lot more exciting than the local business accelerator in a Swedish university and they thought it was a good idea but sort of didn't know how to put together the application.

Pablo Srugo (00:06:47) :
What year was this, this time?

Max Junestrand (00:06:49) :
2023.

Pablo Srugo (00:06:50) :
Okay.

Max Junestrand (00:06:51) :
So May 2023, we make the YC application. We get to interview, we do the interview, we get completely roasted, we get rejected and then the day after we decide, okay, let's take the business accelerator. And let's work really hard, let's reapply in autumn. So we basically strapped in and said, Hey, well, the only way we're going to really learn how to do this is if one, we start to work with some really great lawyers because we're all good engineers. But we sort of don't know our domain and two, that same month, both a company called Case Text got acquired for $650 million by Thomson Reuters. Which is one of the 800 pound gorillas in our industry and I think Harvey announced their seed round or Series A round, Bicycler or OpenAI. Whatever it was and we went, okay, that's interesting. There are other people who think this is a good bet too, but they don't seem to be in Europe. So lets just go and worked extremely hard that summer, super committed. And after a month we signed the biggest law firm in the Nordics, as a design partner and as a client. And they were actually kind enough to let us work from their offices. They're called Mannheimer Svartling and we were part of their incubation and their innovation lab.

Pablo Srugo (00:08:06) :
Were they paying or it was purely just kind of an in kind?

Max Junestrand (00:08:09) :
No, they weren't paying back then, but now they have a sweet deal. But they were critical in the early days and following that, we then got accepted into Y Combinator in August 2023. What was the original question, Pablo? What was the idea? The idea was that LLMs and legal work are going to make sense. So let's run in that general direction, and then we'll tackle problems as we encounter them. And, we'll build a product that compounds the more people use it and the better it gets. And we'll continue to add on more, and more features, and it turns into a suite.

Pablo Srugo (00:08:44) :
Just a few clarifying questions. So the two co founders that are left. First of all, that was the psychologist, and who’s the other in terms of background?

Max Junestrand (00:08:51) :
The engineer.

Pablo Srugo (00:08:52) :
The engineer, okay and how did you know them?

Max Junestrand (00:08:55) :
Oh, we knew each other from a volleyball game. Life is mysterious like that.

Pablo Srugo (00:08:59) :
But then what happened there for them to even reach out to you, or what were you doing at the time, 2021 to 2023? Before starting this?

Max Junestrand (00:09:07) :
So in Sweden, when you go to college, you need to basically go to one college and you pick a program, and then you stay in that program, and you can't do anything else but that program. Unless you basically make an extra application and you need to make it in a funky way. Because you're not allowed to take that many university credits. So I got enrolled in both the business school in Sweden and the engineering university. And so I did those two in parallel, and then I was working full time at McKinsey. I did a bit of venture, I was part of two other startups, but very short stints. Both also in Y Combinator. So I had gotten a lot of exposure to this world and to what it meant to work in fast paced growth startups. And our attitude was pretty much like, how hard can it be?

Pablo Srugo (00:09:56) :
And did they come to you as a VC, or did they just come to you like, hey, we have this project. You should leave the VC thing and come join us?

Max Junestrand (00:10:02) :
Oh, they didn't come to me as a VC. They came to me and said, hey, I was actually working in VC at the time. I was doing due diligence as a contractor for VC. But no, they approached me, and they showed what they had built. And then I thought that looked interesting. And we decided to work on it for that summer, and it turned into something much, much more.

Pablo Srugo (00:10:19) :
When you decide to work on it that summer. Because the high level LLM first law is a very, very big, broad thesis. Which, to be clear, has a lot of credence behind it and obviously a lot of different folks are attacking that, and winning in many different ways. There are so many companies doing well in legal tech, AI for legal tech. But in that summer, where do you decide to focus and why?

Max Junestrand (00:10:39) :
So our initial product that we built was sort of a big dashboard that had six big buttons and it was basically, these are the six use cases that Legora can do. And if I remember them correctly, it was something like ask a legal research question, summarize, check this contract against the policy, and something more, something more. And it was very limiting, but it was pretty good at those six things. And then we, I think, were building with LangChain back then. So we sort of built, you know, small workflows basically that the LLM would go and perform.

Pablo Srugo (00:11:17) :
And that was born out of what? Maybe just to pause there, those six buckets that you decided to focus on. Where did that come from?

Max Junestrand (00:11:23) :
So what I did was I wrote to as many lawyers as I could find on LinkedIn, and I offered to buy them lunch, and to pay their hourly fee to go out for lunch.

Pablo Srugo (00:11:34) :
That's a good offer.

Max Junestrand (00:11:35) :
It was a great offer. They were all too nice to, you know, take my money. We didn't have a lot. Most of them even paid for lunch. But then, I mean, I basically sort of just asked them relentlessly, like, okay, tell me about your work. Tell me about your work and, you know, if you sort of boil it down. I think you get, if you think about the practice of law. It's reviewing, drafting, and researching in some way, shape, or form. And then we sort of took a couple of key tasks from those buckets, and we said, let's build a product around this. But we never really got any usage on that first version because it was summer in Europe, so everybody was away and we were locked in building mode. And then we took the decision to basically scrap the idea of having these skills or buckets or whatever you might call them. And we said, let's go down the sort of conversational AI chat route instead. And then we built this LLM, you know, the agent or conversational AI over the skills. So the skills became sort of tools that the LLM could call on for specific types of tasks that the user was asking them to perform.

Pablo Srugo (00:12:40) :
What drove you to go that way, to go to the chat interface instead of the skills?

Max Junestrand (00:12:45) :
So it basically came down to the idea that the skills would be too limiting and basically meant that for every new use case we wanted to solve for, we would have to build it. Whereas the new models were getting better and better, and better, and the chat kind of lends itself nicely if you want to. If the LLM gets smarter, then suddenly it can do more. Whereas in our case, it wouldn't have. So we wanted to build, you know, software that almost floats like boats and then when the tide rises. All the functionality gets better and what we opted to do instead was we went down the route of, okay, let's have the chat. Let's have the best chat for legal work. But then let's also build other types of UIs or other types of leveraging the models. So we had the chat for basically the first sort of nine months and then we went to what we call tabular review. Which is effectively this very large grid where all the documents get represented as rows and then prompts as columns. So if you put in a thousand agreements manage diligence and you ask it sort of ten questions across the columns. We will run ten thousand parallel API calls and extract all the insights. And that use case or sort of that idea was born out of user frustration that they were using the chat, it was a Spanish word. They were using the chat to do due diligence but it couldn't handle enough context and even now when we have Gemini3 or these bigger models. They sort of get deteriorating quality when you put in too much context and the LLMs get confused. So you really want to break up the task into sort of more bite sized chunks and the tabular view sort of suits itself nicely for that.

Pablo Srugo (00:14:28) :
Well, the other thing about this tabular view and kind of vertical specific UI, is one of the things you have to think about if you're building an AI today is how do you position yourself? So as you kind of mentioned, when the foundational models get better, you get better versus when the foundational models get better, you somehow lose out on a capability. Because now it comes for free or baked in. So when you do a kind of tabular thing that makes sense in this use case, something in ChatGPT, Gemini, they're just not going to do, it doesn't make sense for them. You position yourself closer to that and then when those models get better, every single answer is going to get better than that in that table.

Max Junestrand (00:15:00) :
Exactly, so it's kind of you're rushing towards product market fit. But you're doing it in a way that future proofs yourself and that's also why we're so integrated in Outlook, and we're integrated in Word. And we're integrated in all these other places with all these other systems that lawyers use on a daily basis.

Pablo Srugo (00:15:16) :
Let's dive in a little more specifically to how you structured things with the design partner in the early days. Because these days, beta customers have been a thing forever. Now we call them design partners. Everybody knows you should get design partners if you're building an enterprise. But how you set things up during those first few months matters a great deal in what your product looks like. So how did you do it?

Max Junestrand (00:15:37) :
Well, the first thing was that their head of security came with, you know, a paper stack about this thick and sort of slammed it on our desk, and said, hey, we're not going to work with you until you fill this out. And, you know, we had to get our SOC, we had to get our ISO, we had to get all our ducks in a row. And we actually ended up spending about half the initial angel funding just on security stuff.

Pablo Srugo (00:16:00) :
How much did you raise for Angel?

Max Junestrand (00:16:02) :
Oh, we raised 50,000 euros.

Pablo Srugo (00:16:04) :
Nice, nice.

Max Junestrand (00:16:05) :
That's what we got started on. 

Pablo Srugo (00:16:05) :
Nice.

Max Junestrand (00:16:06) :
We were like noodle, you know, ramen profitability.

Pablo Srugo (00:16:10) :
Yeah, right. Because you weren't paying yourself.

Max Junestrand (00:16:12) :
No, no, no, we weren't. We bought nice screens, that was about it and the other thing we did was we made it very clear that we weren't there just for the office space. We really wanted to work with them and so a law firm is structured like a partnership, right? So you have these partners that are very entrepreneurial and we found the partners in the firm that sort of wanted to go all the way with us. And we sort of forward embedded ourselves a bit into their teams. And in the beginning, it was a lot about user education. I mean, this was pre a lot of people having worked with ChatGPT or pre Microsoft Copilot, right? So this was a lot of people's first experience with generative AI.

Pablo Srugo (00:16:52) :
This is right after, I mean, ChatGPT end of 2022. You're building summer 2023. So it's like six to nine months later, sort of thing.

Max Junestrand (00:16:57) :
Yeah, exactly and all the, yeah, maybe people that used ChatGPT for, you know, writing, helping them write a poem. But they hadn't done anything fundamentally interesting with it and so we have to come in, and also educate them on, hey, if you're working in the banking and finance team in this firm. These are the types of tasks you can do or if you work in real estate, then here's how we can summarize a lease agreement and actually extract the insights, and do all these cool things. The other thing I did was I would bring my computer down to the lunch canteen every day, and I would put it in front of a new person. So I'd be very sort of social and go sit with new people all the time, put it in front of them, and then see their reactions when they first tested the solution. So that was fun and then they came with lots of great feedback. And I think in those early days, it was hard for people to distinguish between, is this like Google or is this like generative AI? You know what I mean? So people went in and they very much thought about, let me ask it very tough legal questions and see if it knows the answer. And it wasn't very good at that in the beginning because the LLMs hadn't been trained on that. A lot of that data, and also the reasoning ability was pretty basic. So we had to sort of steer them to use cases where we knew that they would be successful and then you do things like build prompt libraries, and all these things so that when they get to the system for the first time. They get an experience that we know will be good.

Pablo Srugo (00:18:21) :
You know, summer, fall 2023. You're experimenting, because part of it is the tech is so new. You're not even sure what the capabilities exactly are. What are the boundaries, but where are you trying to get to exactly? What are you trying to be, okay, once we're here, are we going to start selling or are we going to start raising or hiring, or what's kind of the milestone if you want to think of it that way?

Max Junestrand (00:18:41) :
So we got accepted into Y Combinator in August 2023. So that fall was very much about developing the product.

Pablo Srugo (00:18:49) :
Did you move to the Bay area then for a few months?

Max Junestrand (00:18:51) :
No, we went to the Bay area when YC started in January of 2024.

Pablo Srugo (00:18:57) :
Okay.

Max Junestrand (00:18:58) :
But before that we were in Stockholm, we were just hustling from their law firm office, right? And a lot of the things that we were building initially were sort of infrastructure stuff. It was PDF parsing, it was chunking, it was retrieval, it was UI, it was authentication. It was all the basic stuff that you just need to get in order to sell to enterprise clients and the software couldn't go down, right? Your system can't fail and in the beginning, it was also really hard to get rate limits. I remember, and we did not get enough rate limits on GPT on the European servers to serve enough clients. So we were holding off onboarding new clients just so that we would get more rate limits. So all of those problems, you know, you've almost forgotten about them. They seem so basic now, but they were huge back then.

Pablo Srugo (00:19:50) :
Was that frustrating just to be trying to build this new thing, but you got to get all the foundational old stuff done?

Max Junestrand (00:19:57) :
I mean, super frustrating. But now, you know, once it's in place, then you get to do the fun stuff and you get to reimagine how things should work. And I think also what we ran into very early was limitations in the frameworks that were available at the time. So LangChain, for instance, was very popular and it worked to build proof of concepts. But it didn't work to build production grade products. Now it's much better, but I think back then it was still so early. So we ended up building our own agentic framework that we would use for our chatbot. We started getting a lot of interesting inbound because we did a great press release with this big firm that we were working with and almost nothing of that sort converted into any live revenue or deals in 2023. But when we went to Y Combinator in January of 2024. We, one, were in San Francisco, and two, everybody had taken the opportunity to think about it over Christmas and they came back and said, hey, we're ready. We're ready to go.

Pablo Srugo (00:21:03) :
They talk to the rest of their family. They're there, all the kids are saying, here's how AI chatting with you. They're like, oh my God, I'm falling behind.

Max Junestrand (00:21:09) :
I wouldn't think so and we were in San Francisco, and all of our business was in Nordics. I was working between one a.m. and ten a.m. San Francisco time. To take customer calls and our first sales rep, too, was doing the same. And we bought these influencer ring lights to put on our laptops so that it wouldn't look like it was in the middle of the night.

Pablo Srugo (00:21:32) :
When did you sleep? Because I assume then after those all the all the YC stuff like.

Max Junestrand (00:21:35) :
It wasn't very sustainable. I managed to do it for about five days, and then.

Pablo Srugo (00:21:40) :
Okay, yeah

Max Junestrand (00:21:41) :
I tossed in the towel and said the only way we succeed is, if we serve all these clients well and if we get, you know, from 0 to $20 thousand MRR or a million ARR. And I went back, I sort of locked myself in for five weeks, just did customer stuff. I was running around Stockholm with my little briefcase, a lot of travel to Finland, to Denmark, to Norway and we ended the YC batch with almost a million in ARR.

Pablo Srugo (00:22:10) :
Oh, wow.

Max Junestrand (00:22:11) :
With a hundred percent pilot conversion and with zero lost deals versus competitors.

Pablo Srugo (00:22:18) :
Okay, so I've got a bunch of questions about that, but I just want to take you back really quickly because many, many early-stage founders want to get into YC. You have the unique situation where you applied, got rejected, then applied, got accepted. So you've had both the success and the failure path. Tell me a little bit about, not so much the YC process, but what you think the difference was, what that interview is like, what you got maybe wrong the first time or they didn't like the first time, that they did like the second time. Just think of, you know, an early stage founder thinking of applying to YC and maybe what you can tell them that might help them out.

Max Junestrand (00:22:49) :
Well, we changed our team composition the second time we applied and we also had a lot more sort of meat on the bone when it came to our industry. I think the first time we were very nervous, and my co-founder was the CEO and did all the talking. And we had practiced a lot, but we went on. They look for cracks, and there were some cracks. Going into the second interview, we had a term sheet for twice the valuation of the YC terms and I was a bit more aggressive against them. And I basically said, you know, we explained our business. We said what we had done. We said, you know why this is a great opportunity and what we're the team to work on it. But then I was also just like, guys, I give you wanna take this deal, like you're being morons. We're gonna kick ass, and I think they saw the fire and they were like, right, guys are serious, and let's take a bet on us now.

Pablo Srugo (00:23:42) :
So now January 2024 comes around. One question before we get into the sales piece. Which is most interesting, I just want to know. Did you feel, you know, people talk about the SF Bay Area as a mystical place. I spent some time there when I went through 500 Startups ten years ago now. But what was your experience going from Stockholm to the Bay Area, being around all these founders, part of YC Batch. Did you feel a tangible difference?

Max Junestrand (00:24:02) :
Well, we had a lot of imposter syndrome going to San Francisco and, you know, these PhDs from MIT and these XFN and, you know. All these people who look great on paper, and who are very nice people. But then I think what we found was everybody's not as good as they look on LinkedIn.

Pablo Srugo (00:24:22) :
Yes.

Max Junestrand (00:24:23) :
And at the end of the day, you know, if you're successful or not. It's a lot of luck and it's a lot of hard work. And, often the combination of the two. We worked really hard, and we were very focused. And I think we were very intentional about why we were going to YC. I think we had this idea that there were going to be other companies that had come as far as we had, but we, I think, were one of the companies who had come furthest in our batch. We already had a pretty clear product, we had a very clear customer, we had some revenue and for us. It was more about scaling and getting access to the U.S. investor scene, and sort of getting put in that ecosystem. And I think that was really the main value for YC for us. Now, I think you might have a different YC experience if you're more searching for product market fit or searching for what you're going to do and again, I was only there really for the first five days. Every time I've been back, it's excellent. I remember some of the guest speakers and the conversations we had with them. And now I think I'm actually going back as a guest speaker next year myself. You just look at people who are like two, three rounds or years ahead of you, and then you try to copy what they've done.

Pablo Srugo (00:25:37) :
We have tens of thousands of people who have followed the show. Are you one of those people? You want to be part of the group. You want to be a part of those tens of thousands of followers. So hit the follow button. I think, you know, talent is a lot more dispersed than you might think and if you're, you know, high IQ people and you're working hard, and you've got a pretty clear idea of who your ICP is. There's not going to be maybe as much difference as you might think coming from, call it, a tier two,  tier three market to the Bay Area or the people you see around you.

Max Junestrand (00:26:04) :
No, I mean, now, the second biggest office we have is in New York and we're growing like crazy. And I mean, the U.S. is our biggest growth market. And so now we're also making the transition from Sweden to New York, and we're in London and, you know, all these places.

Pablo Srugo (00:26:18) :
So, January 2024, that's when you really start selling. Tell me exactly what the product was at that point?

Max Junestrand (00:26:25) :
Well, we grew from zero to a million ARR, but it was almost all of it was pilot ARR.

Pablo Srugo (00:26:31) :
Okay, but you were. I mean, it's still selling.

Max Junestrand (00:26:33) :
Yeah, we were selling, but it was sort of the initial onboarding of the clients and I think what we experienced then was one, it really takes work to change the way you work.

Pablo Srugo (00:26:44) :
Status quo is tough, man. Status quo is probably the biggest competitor to most startups.

Max Junestrand (00:26:48) :
Right, and lawyers are very, sometimes opinionated about the way they do things. And so showing them how to do something differently is, like, scary. So the change has to happen intrinsically, and we found these people who are, you know, either innovation responsible at the firms or sort of AI native associates who just became champions. And they started showing things that we couldn't even think of, like new use cases. And then that spread in the firm, and then that became a success metric. And we started doing a lot of onboardings on site. We started really thinking about our customer success and post-sales motion. And at some point, we had double the CS reps and legal engineers. Which is what we call our lawyers who come and work with us, to salespeople. So we had a much, much, much bigger post sale motion than new sales.

Pablo Srugo (00:27:40) :
So we'll walk through that, but just to set the context. By the time you started this getting pilots, selling, whatever you want to call it. Motion in early 2024, what exactly was the product? What exactly did it do? How did it work?

Max Junestrand (00:27:51) :
It was a ChatGPT wrapper with really good RAG on your internal documents, as well as externally available legal content.

Pablo Srugo (00:28:02) :
And what would be the main, you know, I'm sure people use it in a variety of different ways. But when you would come and do a demo, what would be the main use cases that you would talk about?

Max Junestrand (00:28:11) :
I would show how I could upload several documents into the chat and then ask Legora to do something, like, hey, write me a two page summary on this, or go and look in this Swedish legal database and search for relevant content and write me a memo on this topic.

Pablo Srugo (00:28:29) :
And when somebody, I would assume, inevitably said, you know, why not just use ChatGPT for this. What would be the answer then?

Max Junestrand (00:28:35) :
Well, ChatGPT enterprise was not available on European servers, so you had no option.

Pablo Srugo (00:28:41) :
You mean from a security perspective? Because you could upload the stuff into ChatGPT consumer, but then you're in a different world of problems.

Max Junestrand (00:28:48) :
Law firms like to follow rules.

Pablo Srugo (00:28:50) :
I would think.

Max Junestrand (00:28:52) :
So look, I mean, the first version of our product was attractive because ChatGPT was not available on European servers.

Pablo Srugo (00:29:02) :
It's crazy to think that, you know, a ChatGPT wrapper is, like, a, it's supposed to be like an insult. But, there are so many great products that get built as a ChatGPT wrapper. I mean, that's what an application is fundamentally.

Max Junestrand (00:29:14) :
In our decans, I do think our document RAG and things like that were much better than ChatGPT. You couldn't even upload a document into ChatGPT at the time. It was really, really simple and when Microsoft Copilot came, a lot of people asked themselves that. They were like, hey, why don't we just buy Copilot? Why should we buy Legora? And then they bought Copilot, and they realized how bad it was, and it just couldn't do anything. And it's pretty decent at, like, summarizing an email, but then transcribing Teams calls is about it. So even with these general LLM products, I always see a need for the vertical interpretation of those capabilities and that, I think, is our job at the end of the day. How do we build the best, most delightful and exciting software for lawyers?

Pablo Srugo (00:30:04) :
So tell me about those three months. I mean, getting from zero to a million in pilot revenue, whatever you want to call it, is a big accomplishment in three months. How do you do that? Mechanically, how do you get leads? How do you close them? What's the ACV? Tell me all those kinds of details.

Max Junestrand (00:30:19) :
Right. So, actually a lot of the leads came from a conference, that I was a presenter at, and they had this startup competition. And the startups were supposed to go up, and pitch themselves. And it was a legal tech-focused conference. And the three startups that went up before me, they presented their company, and they go like, hey, this is our company. This is what we do. This is our investors. This is our fundraising. This is our product. It was really boring and when I went up, I thought, let's not do this. I asked the people in the audience, I was like, how many in here think that there's too much talk and too little action? And everybody was like, yeah, you know, too much talk. So I went, all right, so I'm not going to tell you about our team, and then I had the slide of the team. I'm not going to tell you about our marketing, and then I had the marketing, and I was not going to tell you about our investors. I'm just going to show you.

Pablo Srugo (00:31:08) :
And this was off the cuff. You hadn't planned. You just kind of decided.

Max Junestrand (00:31:11) :
I hadn't planned it. No. So I went over to the computer, and I tabbed out from the PowerPoint. I go to Google, I log in on Leya, Leya back then not Legora and I'm shaking. Because I've never done a demo for two hundred people before and our product failed a lot in those early days. So it was kind of a fifty-fifty if it was going to throw an error, I felt like and I go up, I show a couple of use cases, and everything works perfectly.

Pablo Srugo (00:31:39) :
Wow.

Max Junestrand (00:31:40) :
And the crowd is like, well, that was awesome. And then I'm like, ah, quick. So I go to our webpage, and I'm like, and here's the book a demo. And then my phone starts to vibrate, because I get all these demo bookings. So I actually did one hundred fifty demos off the back of that presentation.

Pablo Srugo (00:31:58) :
Wow, that might be the most successful conference of an early stage startup ever. You might have set a record there.

Max Junestrand (00:32:05) :
But I did something different. I did something that they would find exciting because everybody was talking about AI, but nobody showed them anything and I sort of went, well, here it is. It's the sort of, you know, take the clothes off the emperor and after that. I think we had solidified ourselves as the Nordic legal AI company and a lot of those demos were then the ones that actually became customers. And I remember our biggest deal that we signed during YC, and I was sitting, it was like three a.m. I had just showered, and I was sitting by our table, and I was trying to talk not too loudly. Because everybody else was sleeping in the room, in the apartment and it was $45k. And I was like, this is amazing. This is the biggest deal ever and the team was wild, and we celebrated. And then they kept coming in. Another $30k deal, another $20k deal, another $40k deal.

Pablo Srugo (00:32:58) :
What was like your demo to close ratio back then? Roughly, like was it 1 in 3, 1 in 4, 1 in 2?

Max Junestrand (00:33:03) :
No, I had like fifty five percent.

Pablo Srugo (00:33:07) :
Wow. That's when you know, man.

Max Junestrand (00:33:09) :
We were using Atio at the time as our CRM and that sort of gave us the win rates.

Pablo Srugo (00:33:14) :
I was speaking, by the way, to another founder the other day. Who also had an incredible demo to close rate and he said to me, it was so good that when a demo would come in, I just assumed it would close. I would just bank that ARR because I'm like, it's going to happen. Which is just a crazy place to be in.

Max Junestrand (00:33:30) :
Yeah, it's wild. But the pipeline had begun from November. So it was a three month sales cycle.

Pablo Srugo (00:33:37) :
And how did you structure the sale? Did you do it as a pilot, or did you do annual contracts? Was it?

Max Junestrand (00:33:42) :
We did free pilots, I think, oh, no we did paid pilots mostly. So it was a paid pilot, two week or four week. I would onboard everyone. I would work with them one on one and then almost all of it would converge. We had one hundred percent conversion rates back then.

Pablo Srugo (00:33:56) :
And how did you charge? Was it per seat? Per seat per month? Is that the model?

Max Junestrand (00:34:00) :
Yeah, per seat. So after that, I mean, this is the interesting part. So after that, we raised from Benchmark.

Pablo Srugo (00:34:06) :
Which, obviously, I would assume. But I was going to ask, how did you raise $10 million from Netron? But I assume they saw the hockey stick, you know, they got the thesis, and the rest is history, sort of thing.

Max Junestrand (00:34:15) :
You met me now, right?

Pablo Srugo (00:34:16) :
Right, right. They met you. Yeah, there you go.

Max Junestrand (00:34:19) :
I mean, from what I've heard from Chetan and Peter, we had the meeting. It was thirty minutes. I walked out the door. They then walked to their office, and Peter turns to Chetan and goes, you know, I've always liked the Swedes, and Chetan goes, yeah, and we're doing this deal, right? And Peter goes, fuck yeah. And then they did the deal.

Pablo Srugo (00:34:43) :
That's it.

Max Junestrand (00:34:44) :
That was the ICM benchmark.

Pablo Srugo (00:34:46) :
I mean, that's how I see the seed really should be, to be honest.

Max Junestrand (00:34:49) :
But yeah, I mean, look, I think they spent a lot of time with some other companies in our industry, and they were maybe not as excited about those.

Pablo Srugo (00:34:57) :
Well, you build it. That's the point. You build this pattern recognition by talking to a lot of founders. When the right founder with the right idea shows up, it really shouldn't be about two weeks, three weeks of diligence. I mean, you kind of know at the time there's a bet worth taking, and many won't work out, but it's a bet worth taking. That's really that.

Max Junestrand (00:35:13) :
And almost immediately after, we got a preempted Series A term sheet. And, you know, we were sort of debating if we should take it or not. What we landed on was there was probably an opportunity to build a really big European company and focus on that. But we wanted to build a global company, and that meant going to the States. And so raising that capital quicker propelled that timeline. And then came probably the most important decision in the company's history. Which was in our first board meeting in May of 2024. I sat down with Benchmark and Redpoint, and I told them that we're not going to sell for the next six months because the product was not ready. I thought we should sort of go into general availability. So following all of those pilots, we turned around and just fully focused on not churning any of them and making sure that they got a great experience. Because the product had some scalability issues and it was quickly put together and so had to be reworked.

Pablo Srugo (00:36:22) :
What were the signs, by the way? Was it qualitative, people telling you stuff, or did you see things that made you say, no, we need to spend some time on this?

Max Junestrand (00:36:28) :
You know, you run into rate limit issues where you're, like, tossing a thousand documents and starting a big review. And then suddenly the system breaks. For me, it was trying to do something similar. So you need to build a good queue system, and then you need to build a good file processor, and you need to go and build all these things.

Pablo Srugo (00:36:45) :
That's the no brainer stuff. What about more of the feature piece, like everything else?

Max Junestrand (00:36:49) :
The features almost did not develop between May and October. It was like, they're the same product.

Pablo Srugo (00:36:56) :
Really, just usability and scalability.

Max Junestrand (00:36:58) :
We tried one or two features during the summer. Because people were bored of just doing infra. But both those products were really stupid, and we shut down those products.

Pablo Srugo (00:37:08) :
What did the VCs think after putting, I mean, combined, what is that, $35 million in your company? You're like, okay, for six months, we're just building.

Max Junestrand (00:37:13) :
Yeah, I don't think they were thrilled, honestly and then we're like, all right, we'll just give this twenty five year old $35 million.

Pablo Srugo (00:37:20) :
Well, that's part of the thing, right? If you were a proven founder, you'd build and exit a company. You could show up to that meeting and say exactly what you said and be like, yeah, of course, Max, no brainer. You want to build for six months, it's a no brainer thing. But you're twenty five years old, man. This is your first thing. You just took a bunch of money to grow, and you're like, I'm building. Yeah, that's interesting.

Max Junestrand (00:37:36) :
Well, look, you know, I felt confident that if we were onboarding all these very prestigious and nice firms on our products, and it didn't work. Then we'd burn our shot, and what has happened is that a few other companies burned their shot too. And now, when we show up, it's like, this is ten times better. We should switch and it's like, yeah, you should. Let's do it.

Pablo Srugo (00:38:00) :
Did you compete with Harvey or some of these other platforms for legal? Or is there a way for people to have both? How does it work?

Max Junestrand (00:38:07) :
No, I think we compete pretty fiercely. There's some alternatives. I think it's becoming clearer by the day who's building for the long term and whose product advantage is compounding. And I don't wake up every day thinking about them. We don't even talk about them that much. We wake up, and we set the roadmap, and then other companies copy that roadmap. Which is great because that means that we can start building the next thing and we can always make sure that our interpretation of the model capabilities or new pieces of the stack are what we should go and build because our clients are asking for it. We just focus on that.

Pablo Srugo (00:38:47) :
So that's today, but during those six months when your head's down building and I'm sure Harvey's raising rounds, and other people are raising rounds. And it seems like they're accelerating, are you just as comfortable, or are you kind of having sleepless nights about it?

Max Junestrand (00:38:57) :
We've never been very nervous about what's happening in the market. Look, Thomson Reuters bought a company for $650 million back in 2023, and we've churned that piece of software from many accounts that prefer us.

Pablo Srugo (00:39:16) :
But what's the, maybe just walk me through, again, it's one thing now where you've proven it out. But in those six months while you're building, even then you churned?

Max Junestrand (00:39:23) :
I mean, there was nothing to discuss. The product had to scale.

Pablo Srugo (00:39:26) :
Right, it had to be done. There was no other alternative.

Max Junestrand (00:39:28) :
I mean, what do you want to do? Do you want to show up and it doesn't work? No, it has to work, you got to fix it and we fixed it. And then we launched in GA in October.

Pablo Srugo (00:39:38) :
How was that launch? When you say launch, what did you do?

Max Junestrand (00:39:40) :
Well, I mean, the launch was really, we started onboarding clients again and so all of those customers from back then. They can sort of tie their ears a little bit. We told everybody that we were too busy with onboarding new clients and we told everybody, oh, you know, it's summer. It's so hectic and I, you know how it is in Europe. We're all going on a two month vacation, but there was no vacation that summer.

Pablo Srugo (00:40:04) :
Were you taking demos, closing them, and just putting them on the side, or how were you dealing with it?

Max Junestrand (00:40:10) :
Oh yeah and I just said, oh, it's amazing that you guys want to come on Legora. We really want to serve you, but I hate to tell you, I mean, there's too much demand. And my team just cannot physically onboard you before October. There's no way, so we'll put you in the queue.

Pablo Srugo (00:40:26) :
Did that almost help? It could have a pervert, a kind of counter effect where it's like, oh my God, it must be so great.

Max Junestrand (00:40:32) :
Of course and, you know, we wouldn't account for any revenue, and they could still cancel the contract if they wanted to. We were very easygoing back in those days.

Pablo Srugo (00:40:40) :
But to be clear, and that's obviously the way you have to do it. There is kind of this psychology of once you've signed a contract and you're waiting for the thing. The last thing you actually want to do is cancel it. Because you're excited, you want to go forward, the momentum is let's go.

Max Junestrand (00:40:52) :
And look, I mean, the reality is a lot of the firms are on vacation, and you cannot onboard a new piece of software in Europe between June and September, end of August.

Pablo Srugo (00:41:04) :
I don't know if you remember this, but when you do go live in October. How much, let's say, booked ARR do you have just sitting there to onboard? What are we talking about? A million, half a million, $5 million? Do you know what I mean?

Max Junestrand (00:41:18) :
We've doubled revenue every quarter since October, and we still continue to do that. Yeah, we were north of what we did in YC.

Pablo Srugo (00:41:29) :
When you go into October, you might have had a $2 million signed contract. You just need to onboard them, something like that. Round numbers.

Max Junestrand (00:41:35) :
Yeah, exactly and the important thing for us was we need to build out the capability to be with our clients on site in the U.S., in the U.K., in France, in Spain, in Germany, in Australia. We've scaled the organization very, very quickly now. Because it's not enough to just call them and go, hey, do you guys want the software? It's no, it's an extraordinary amount of change management, and we've built out the capabilities in house to help manage this. So sometimes they refer to themselves as recovering lawyers. We like to call them legal engineers, and it's really sort of tech savvy lawyers that we forward deploy with our clients.

Pablo Srugo (00:42:17) :
I'm going to talk about the fulfillment side of it in a second. I just want to ask quickly about top of funnel, right? Leads, whatever. The demo to close obviously is outstanding, so that very much helps, but how are you refilling that funnel? You know, we're talking, let's say for the last twelve months, what's your main channel go to market.

Max Junestrand (00:42:33) :
A lot of it is word of mouth. A lot of it is they see us at an event, or they see a competitive firm going with us and they go, oh, what's that? That looks interesting. I think we've also clearly taken the sort of pole position in our market in a lot of ways. Where we're actually educating them about how they should be thinking about applying AI in their business. Because it's not just to go and buy the tool and then you're done.

Pablo Srugo (00:43:03) :
What percent of sales are inbound versus outbound?

Max Junestrand (00:43:05) :
Oh, I think it's probably. I actually don't know that number anymore.

Pablo Srugo (00:43:10) :
I should, but is it more inbound than outbound or like 50/50-ish?

Max Junestrand (00:43:13) :
More inbound for sure.

Pablo Srugo (00:43:14) :
But you are running some outbound? Because I was thinking maybe it was all inbound.

Max Junestrand (00:43:18) :
No, no, no, almost all inbound. I think it's like eighty percent inbound. We have a small team of SDRs.

Pablo Srugo (00:43:24) :
Yeah, because I mean, you're tapping at this point, I would assume most law firms are thinking about how to implement AI in their organization. They're having these discussions. It's more about who are we going to go with? When are we going to do it? Versus, oh, I didn't know I could have AI in my law firm, right? Which is a lot of times when you go out, that's what you're doing.

Max Junestrand (00:43:39) :
Exactly, so you want to be top of mind and I think we've also, we did a name change, right? At the beginning of this year.

Pablo Srugo (00:43:46) :
Why'd you do that?

Max Junestrand (00:43:47) :
There was another company that thought that we should change our name.

Pablo Srugo (00:43:50) :
I see, okay.

Max Junestrand (00:43:51) :
We didn't have a lot of things to put up against that. We took that opportunity to almost relaunch, and I think we did that relaunch very well and we combined it with entering the U.S. So we had just signed Goodwin and Procter as our first U.S. client. Which is a massive, massive firm and their brand halo, I think, also sort of went a bit on us as well.

Pablo Srugo (00:44:17) :
Tell me a little bit about the fulfillment side. You've gone from, how many people were you at, let's say, the end of '24?

Max Junestrand (00:44:22) :
Forty.

Pablo Srugo (00:44:23) :
And now you're?

Max Junestrand (00:44:24) :
Two hundred and twenty.

Pablo Srugo (00:44:25) :
So this is the part that's hard to comprehend, right? You go from forty to two hundred and twenty, it's just so many people that are added on a weekly basis, and the rate of growth for the organization. How do you think about, because this is what happens post insane product market fit. This is the effect of product market fit. How have you structured that? How do you think about everything that that entails, both on the culture and employee piece, and then there's the customer support, customer success piece of that as well?

Max Junestrand (00:44:55) :
So I think since January, it was pretty clear that this was going to be an exceptional year and we never made yearly plans. So we always plan quarterly and since we did so well every quarter, we were like, holy shit, we need to scale to meet this demand. And we ended up hiring a lot in the first half of the year. And everybody always comes to Stockholm to get onboarded. But then after summer was when we had a really big batch. So I think August fourth, it was like forty people joining. Which was like a thirty percent increase in the business and then since then, it's been about twenty new joiners every two weeks or every four weeks. So we have an amazing person. Her name is Michaela. She runs all our onboarding, and we sort of have them interact both together and with all the other teams. So they get to see how engineering works. They get to see how marketing works. They get to see how finance works and they get to learn our product, and they get to learn our customer. And very often will bring a customer to do a session even for the new batch. So we take that very seriously, and the hardest thing is to hire at their pace, right? Because if you get the wrong people in, even at the time when you're two hundred, it sort of becomes like a disease, and it spreads. And you need to be ruthless about getting the right people in, having a very good interviewing process or selection process and then also setting very clear expectations when people join the team. That this is a rocket ship, and we are working for a generational opportunity that is very rare to have the opportunity to do. And so we're going to run pretty hard at it.

Pablo Srugo (00:46:36) :
I know you don't think much about competition, but just tell me one thing. When you do go into a sale, especially with these enterprise clients. It's got to be a bake off, I assume in many cases, it's got to be some sort of competitors on the table. What do you think is the key to beating competition when you're selling to enterprise?

Max Junestrand (00:46:50) :
Well, to be clear, a lot of those A B tests are also ones that we encourage. We're like, you should test all of us against each other and pick the one you want. I've found the selection process to basically fall on three things. Current capabilities and how good your product is, and how much they can achieve with it. Two, the team that they got to work with, how much that team cares, how much they show up, how much effort they put in and thirdly, what does the roadmap look like, and how is this company going to partner with you for the long run to make sure that you are successful. And it's a combination of all those three things. And the reason why the last is important. The third bucket, is every quarter, the AI landscape changes a little bit and so you actually want to bet on a team that is going to put your interests first. And who is going to work to incorporate all of the latest developments for you, and also bring you into that conversation. So that you, as the end practitioners, get to be part of the future of your business. Because AI is reshaping and fundamentally rewiring how so much business gets done. And the clients, the law firms, the lawyers, they want to be part of that conversation. And when we win, which is a very high percentage. At least the one we track in Salesforce. I find them very much picking on the second and third ones. Because if you just look at the products, I'm of course biased, but I think that we've actually secured quite a significant product lead. And I also think that product lead compounds very quickly. One of the reasons that we've been able to do this is we're a very, very engineering driven company and engineering driven culture. We're fully in office, everybody's shipping really fast, reiterating really fast, and we're also not afraid to take big bets on what the next set of functionality will be. Give me one example. We just released something we call the portal, which is basically a way for law firms to productize their knowledge and then share that with their clients in a white labeled way. That has never happened before. It's never been possible, but now they can monetize that value and push it directly to clients. And it's really fun to go up into A B tests or A B C tests because culturally, we're very competitive, and we care about being the best, and we can just run really, really fast. It creates a culture of winning. Everybody wants to be part of a winning culture.

Pablo Srugo (00:49:31) :
How often would you say you beat competitors when you go up against them? Do you know your win rate?

Max Junestrand (00:49:34) :
Yeah, it's about eighty five percent.

Pablo Srugo (00:49:38) :
Nice, that's crazy, man and last question on that before we take it to the final three questions. How long were you leading sales for? When did you kind of pull out and start putting other people in charge of sales?

Max Junestrand (00:49:48) :
So I ran sales until we were eight reps.

Pablo Srugo (00:49:52) :
Perfect. Well, listen, let me stop it there and let me ask the kind of three questions we end on. Today, you've grown from zero to tens of millions in ARR in a year. So you have insane product market fit. But when was the moment when you personally felt like you had found true product market fit?

Max Junestrand (00:50:05) :
I think I felt a fake product market fit when I had all those demos come in after the conference. Because a lot of those demos then turned to contracts, but it was not until our product was really scalable that it worked. But I knew that there was demand for it then. So then I was like, all right, we're on the right track. Now, when you have thousands of lawyers coming onto the platform every day, every week, that feels like product market fit, but in a different way. I think following our sort of GA move and seeing that it worked. That was probably the moment I went, all right, this is it. Now I feel very confident.

Pablo Srugo (00:50:43) :
Was there ever a time when you thought things actually would not work out and maybe Legora would just fail?

Max Junestrand (00:50:47) :
Oh, I mean that first summer, things were really sort of up and down. And I wasn't convinced back then that following the failed YC interview, following that everybody was on vacation and nobody wanted to talk to us. I was like, is this really the bet to make? But then when the design partner came back and we started working with them, and it got energizing again. Then just that, I've never looked back.

Pablo Srugo (00:51:15) :
And then last question, what would be your top advice for an early stage founder who is trying to find product market fit?

Max Junestrand (00:51:20) :
Don't outsource the sales and practice good presentation skills, and storytelling. And these, you know, I'm an engineer. I had to learn this. It doesn't come super naturally to me, actually. I had to train a lot to get to where I am on that and I think it's a very underappreciated or undervalued skill that founders need to have if they want to succeed.

Pablo Srugo (00:51:46) :
Perfect. Well, Max, thanks so much for jumping on the show, man. It's been great.

Max Junestrand (00:51:49) :
Thank you, Pablo.

Pablo Srugo (00:51:51) :
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.