He "kind of" had PMF for 8 years—until, after a rebuild, he raised $100M | Ben Alarie, Founder of Blue J

Ben Alarie spent 8 years building Blue J with "partial product market fit"—real customers, real revenue, but no real market pull. Then he made a bet that would either kill the company or 10x it: he put the existing product in maintenance mode and gave his team 6 months to rebuild everything from scratch using a technology that barely worked.

Two years later, Blue J went from $2M to $25M in ARR. They're adding 10 new customers every single day. NPS went from 20 to 84.

This isn't a story about getting lucky. It's about a founder who knew—with absolute conviction—that the market would eventually arrive, and made sure he was ready when it did. But it's also about the danger of fooling yourself into thinking you have PMF when you only "kind of have PMF."

Why You Should Listen:

• Learn the brutal difference between fake and real PMF
• Discover when to abandon millions in existing ARR to go all-in on something else
• Why "time to value" might be the single most important metric for word-of-mouth.
• See what it takes to survive until the market is ready.

Keywords:

startup podcast, startup podcast for founders, product market fit, founder journey, early stage startup, startup pivot, AI startup, SaaS growth, founder advice, hypergrowth startup

Chapters:

(00:02:00) Starting BlueJ
(00:9:26) Introducing AI to Tax Research
(00:12:44) Starting to Build
(00:17:03) Not Having True PMF
(00:19:44) Believing in Retrieval Augmented Generation
(00:25:34) Updating to V2 of BlueJ
(00:30:58) The Necessity of Time to Value
(00:33:47) When You Knew You Have PMF
(00:38:19) One Piece of Advice

Chapters

• 2:00 Starting BlueJ
• 9:26 Introducing AI to Tax Research
• 12:44 Starting to Build
• 17:03 Not Having True PMF
• 19:44 Believing in Retrieval Augmented Generation
• 25:34 Updating to V2 of BlueJ
• 30:58 The Necessity of Time to Value
• 33:47 When You Knew You Have PMF
• 38:19 One Piece of Advice

Transcript

Ben Alaire (00:00:00):
What we learned was there was partial product market fit. The models that we built were actually functionally very useful, when you had one of those kinds of cases. Because they would give you directionally a really good sense of how to handicap your odds if something were to go into litigation or be challenged by the government. And he said, this is a tricky problem. It took us two weeks to do this. Internally, can you try this? And he described the question and I push enter. I'm like, OK, let's see what BlueJ comes up with and BlueJ started bringing in the answer. This guy stood up, walked up to the screen and he was reading it line by line. And he said, that's the answer we came up with. It was fully conversational, you didn't have to wait nearly as long for an answer, it was like 15 seconds rather than 90 seconds and we grew from like roughly $2 million in ARR to just shy of $9 million in ARR by the end of 2024. It's like, OK, now this is really working.

Previous Guests (00:00:57):
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 (00:01:09):
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 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. Ben, welcome to the show, man.

Ben Alaire (00:01:24):
Thanks, Pablo.

Pablo Srugo (00:01:25):
Most people I interview on the show, I just meet right before I interview them. We've been investors in your company for a while now. I think we invested in like 2017, 2018, and you've been working on this since 2015. But you hit a major inflection point. When would you say? Was that a year and a half ago, two years, is that right?

Ben Alaire (00:01:41):
Yeah, I would say about two years ago.

Pablo Srugo (00:01:43):
And you just raised, what was the latest round?

Ben Alaire (00:01:45):
We raised a Series D round and that closed in July of 2025, and it was $122 million USD.

Pablo Srugo (00:01:53):
So let's start at the beginning, right? You started this in 2015. What were you doing at that time? And what did you see that made you start BlueJ in the first place?

Ben Alaire (00:02:00):
You know, I started out as a tax law professor. I'm still a tax law professor at the University of Toronto. But I had gone to undergrad and studied economics, finance, philosophy, went into law school at the University of Toronto. Did a graduate degree in economics concurrently with my law degree, and then went on to graduate studies in law. Law school and then I moved to Ottawa, and I was a law clerk at the Supreme Court for a year. And immediately after that, I became a tenure track law professor at the University of Toronto at age 26. I think maybe the youngest.

Ben Alaire (00:02:34):
How does that happen? Cause that's not, yeah, no, no.

Ben Alaire (00:02:36):
Maybe the youngest tenure track law prof in the history of the law school. I think a whole bunch of things went really my way on that one. I had an interesting research agenda, I had published some articles as a law student and I was super keen on the academic path. I was really excited about the life of the mind and doing research, and going deep on tax law. And so that's what I spent the ensuing decade really focused on. So from 2004 right through to when I started BlueJ, I was full tilt into becoming the best possible tax law academic that I could be and so I wrote dozens of academic articles. I co-authored several editions of the leading tax book, used to teach tax in Canadian law schools. The book's called Canadian Income Tax Law. I co-authored with the lead author, David Duff, the second, third, fourth, fifth, and sixth editions of that book.

Pablo Srugo (00:03:32):
Did you like it? Was it everything you kind of expected it to be? The professor life?

Ben Alaire (00:03:36):
The professor life is an amazing life. You get to research things that you're interested in. The students are evergreen and very talented, and very ambitious, and the classroom discussion is always interesting and engaging, and challenging. And if you approach law teaching with the right attitude, you learn as much from your students as they learn from you. And so, yes, being a law professor is an amazing privilege. I really loved it, and I still do love it. I may return to it post-BlueJ life on a full-time basis. But it was in the context of actually becoming associate dean of the law school and leading a reform of the curriculum that got me thinking about AI in the first place. And so I petitioned the dean. I said, if I'm going to become associate dean. One of the things I'd like to do is to lead reform of the JD curriculum, because I had just been through the law school curriculum from 1999 to 2002. When I was a law student and I felt like it could be done in a different and probably a more effective way.

Pablo Srugo (00:04:42) :
Was it just outdated? I mean, a lot of education, I think, is outdated. Was that kind of just the general?

Ben Alaire (00:04:46):
One of the characteristic things about the curriculum at the time was, it was evaluated through virtually entirely a hundred percent final examination. So you would start the first year of law school in September and you would have virtually no evaluations accounted for anything until April. And then you'd set a series of six or seven final exams, each of which would be worth a hundred percent of the final grade.

Pablo Srugo(00:05:06):
Wow.

Ben Alaire (00:05:07):
And so you wouldn't really know how well you were doing until you got your grades sometime in the summer, following your first year of law school. Which seemed like not the best way to allow people to learn and iterate as to their learning approaches and really figure out how well they are learning the law. And so it was as a result of co-leading this reform that I started thinking about AI and it happened pretty innocently. I asked one of my colleagues what they thought about the curriculum reform and they said, Ben, you really shouldn't do this. I said, oh, why? They said, well, the last time we had a major reform of the law school curriculum was in the early 70s. 1970s, and it led to some hard feelings among some faculty who never really subsequently got along. After that effort as well as they did before the effort. They said it can just people feel strongly about the curriculum and how they approach it. And so it's a bit of a minefield, and you'd be better off just focusing on your scholarship. And I didn't take the lesson from that, that I think that your colleague who is hoping to be helpful was wanting to deliver. Which is like, just kind of back off and so I thought, oh, that's really interesting. It's been forty years since somebody tried to do this. I thought, okay, for one thing, it's probably far overdue. We should do it again and move to change the curriculum. How we're approaching teaching the law and the second thing I took away from it was this observation. Which was, well, if it took forty years for someone like me to show up and suggest that we should really do a deep dive into changing the curriculum. It may be another forty years before someone like me shows up and wants to do it again. And if that's true, and I had this thought. If that's true, then what does that mean? And I immediately started thinking about, well, AI and machine learning.

Pablo Srugo (00:06:53):
Just trying to go to the future, not the present but kind of like what's coming next.

Ben Alaire (00:06:57):
Yeah, let's imagine what is kind of the societal backdrop forty years from now. So at that point, it would have been like thinking about the year 2050. What's probably going to be true in the year 2050, that we should be taking into account? If we're going to rebuild this curriculum, what should we be planning for to make it robust to? And I was like, well, Moore's law has been pretty robust and computing power is doubling every couple of years. And its affordability is increasing, you know, improving exponentially as well. And it's like, okay, so that's coming. And then I thought about the fact that legal information is even at that time was virtually all digital, and algorithms were improving significantly. So I was aware of some of the work that Geoff Hinton and other colleagues were doing in the computer science department at the university of Toronto and I kind of felt like, oh man there is a free train with AI on the front of it coming directly for the law. And wow, this has big implications for this curriculum reform but also for the legal system generally, for the judiciary, and also selfishly, and closer to home, like for me personally. And so I started to think about, okay, what if I'm standing at the front of a law school classroom in the year 2040. Thinking, oh man, I saw this coming in 2012, 2013, 2014, and I decided not to change my professional trajectory. And I just decided to stay on the same path. And I started to think I would probably feel some profound regret about that, if I weren't to do anything about it. And then I thought, well, what is it that I should really do about it? I mean, OK, changing the curriculum is one thing and we'll work on that, and introduce the curriculum reform. And I think it's been successful, you know, we have another more than a decade of experience with the reform curriculum, and I think it's been successful. But I started thinking, OK, so what does this mean for tax law? And I'd been through this exercise of revising several editions of this tax book. It's like a 1500 page introductory text on Canadian income tax law and it's such a manual job to update it. There are new cases coming down from the tax court, from the federal court all the time, the Supreme Court. The legislation is being frequently amended by parliament. Things are always changing in tax law and so you have to go through, and you have to laboriously research each point to see what has happened. New and it's always true that no sooner is that new edition of the book published, than the laws change and it's already out of date.

Pablo Srugo (00:09:21):
It's outdated.

Ben Alaire (00:09:23):
There's got to be a better way to do this. I was like, okay, well so clearly there's a role for AI in improving tax research. Okay, so what does that mean for me? I was like, well, somebody's going to improve this. I started to think, well, who's that going to be? I started to wonder, well, might it be me? Surely not, surely it's not a law professor who's going to be leading this effort. Surely it's going to be somebody else. But I thought about the different classes of people that it might be. It might be a tax law partner in a big firm, but that's so extraordinarily unlikely. Because they already have clients that they're servicing, they're part of a partnership, they're kind of bought in, they kind of have dedicated their lives to that path. It could be somebody who's doing tax at an accounting firm, but it's a very similar. It's a professional services role and they're on that path. They have clients, they enjoy a relatively good living and what draws somebody into that kind of career path is it's a known path, and it's relatively predictable. And they are not dispositionally typically going to be the kinds of people who are like, oh, I know what I'm going to do. I'm going to do a tech startup and kind of roll the dice, and see if they can make it work. Because the opportunity cost of doing that would be so significant. So I thought, well, OK, it's not going to be somebody with a tax law background. Might it be somebody who's more kind of pure tech, like an AI or machine learning background? And again, I thought, well, probably not, because they would need to know a lot more tax law than, they do. It'd take them a decade to catch up to where I was in my understanding of tax law, and probably they're not just not all that interested in learning tax law. And they wouldn't really understand the pain of tax research, right? They probably just prefer to go work for one of the FAANG companies or do a startup in some other area, and so it's probably not going to be one of those folks. It's like, well, maybe it is. Maybe it could be me and so talking to a number of my colleagues at the law school. And a couple of really impressive people, Anthony Niblett and Albert Yoon. Anthony Niblett has business and law degrees from the University of Melbourne. He grew up in Australia, a PhD in economics from Harvard and taught at the University of Chicago Law School for a number of years before joining the faculty at the University of Toronto. And Albert Yoon, who grew up in upstate New York, went to Yale College and then went to Stanford. Did his law degree at Stanford, his PhD at Stanford, and then taught at Northwestern for several years before joining the faculty. And we frequently have lunch together and talk about ideas and say, guys, like this is a strange idea. But I think we should start a tech company to leverage AI to do tax research. Both of them were like, OK, Ben, we're on board but you have to take the lead. I'm like, OK, that's fine. You guys are really impressive, brilliant. I can't imagine having better co-founders in terms of intellectual capabilities and being really great people to spend a lot of time with. So we started BlueJ, and it happened to coincide with the end of my term as associate dean. Which was great, because I was entitled to a year of administrative leave. Essentially a sabbatical after that and so I could like just throw myself into this for a year, on a risk free basis kind of, Pablo. So I was like, this is great, because I don't have a huge opportunity cost. I have this time that I can dedicate to doing whatever it takes to get this off the ground.

Pablo Srugo (00:12:37):
Did you know what you wanted to build? Or you just knew it was AI and tax law. And you just had to explore the space to figure out what you might build?

Ben Alaire (00:12:44):
What I wanted to build from the outset is what we've now, eventually been able to produce.

Pablo Srugo (00:12:48):
Ten years later, yeah.

Ben Alaire (00:12:50):
So it took ten years, but what I originally wanted to build out of the gate is a system that could automate all of tax law research. So you could just type any arbitrary tax research question into BlueJ and BlueJ would just produce a world-class answer leveraging primary sources of law. So like, legislation and case law, and maybe commentary, academic commentary, professional commentary on the topic. And give you essentially a full legal memo, a takedown of that particular question. That was utterly science fiction in 2015, when we started. So we had to start with what we could actually get to work reliably, and what it turned out, we could get to work reliably was supervised machine learning models that could predict how courts would resolve certain kinds of tax questions. And so there are a number of these examples throughout income tax law. One example is whether a worker is an independent contractor or an employee for tax purposes. That's a nice conventional example and there have been hundreds, and hundreds of these cases decided by the courts over the years. And it's really tricky, because you could have an agreement that says you're an independent contractor, and there's not an employer-employee relationship here. And the courts will ignore that agreement, and say, actually, for tax purposes. This worker really is an employee because they'll look at the functional relationship between the hirer and the worker. And say, no, there's too much control exerted over the worker's work by the hirer here for us to ignore that. And say that this person really is an independent contractor, and so they'll reclassify that relationship. And that has big consequences, because then the employer of the deemed employer should have been withholding income tax at source. Should have been withholding EI and CPP contributions. Payroll taxes on that worker's earnings and should be paying the CPP and EI matching contributions, the employer matching contributions. So they're like really draconian tax consequences that flow from that reclassification of that relationship. That's one example, so we built a bunch of these different predictive models. Which performed extremely well, better than we thought they could perform and so we're predicting with better than ninety percent accuracy how courts would resolve these kinds of cases.

Pablo Srugo (00:15:06):
And you could do this at cross countries? Or how did you have to focus in on specific states?

Ben Alaire (00:15:11):
We started in Canada. So it's Canadian federal income tax law. So it's essentially the same across the country. There are some complications with Quebec, which uses the civil law rather than the common law. But it turns out that those cases are just as predictable as they are in the common law provinces as well, and so we were off to the races. But we learned and I think this is key to the topic of the show. What we learned was, there was partial product market fit. The models that we built were actually functionally very useful when you had one of those kinds of cases that you were analyzing either as a tax lawyer, as a tax accountant, or even in government. Those models were very good, because they would give you directionally a really good sense of how to handicap your odds if something were to go into litigation or be challenged by the government.

Pablo Srugo (00:15:56):
Who are the customers, by the way? Just as a context, you're selling this to governments or you're selling this to law firms?

Ben Alaire (00:16:00):
Yeah, law firms, accounting firms, and in government. Had some conversations with corporations. That was a tougher sell for the tax product because they just really didn't tend to have the same frequency of tax work. The biggest limitation of V1 of BlueJ, is that we only had these issue by issue models and so they were really good at what they did. But it didn't satisfy the original vision. Which was something that could handle any tax research question that you threw at this thing, and so we grew it. It was encouraging, the progress was encouraging. This is why Mistral invested. It's why we were able to raise a Pre-Seed round, a Seed round, a Series A, a Series B.

Pablo Srugo (00:16:38):
Walk me through, because this is really important from a position of not having full product market fit. Which is you got to a few million in ARR at a good pace. As a founder that never done this before, you would think, OK, if I can get a few million ARR with no resources. Then for sure, if I raise more and I've got more stuff going. I'll get to the next phase. That's got to be easier and sometimes it's not, right? And it kind of happened to you for a while there. Just describe maybe why that is.

Ben Alaire (00:17:02):
Technically, what we had built was world class and we could demonstrate it, and people could test it, and they would trial it. And they'd be like, yeah, this is this technology does what BlueJ is promising it can do. It's actually predicting how these cases are going to go and it's fantastic at that. What was less appreciated is the behavioral change necessary to adopt that and to actually get the value out of the platform. And so one of the big problems with not having a solution that could answer any tax research question is the following would happen. And this is a stylized version of what would happen but people would use it for something that BlueJ does, and they get excited. They're like, oh, this is really great and then that issue would come up again. They'd log in, they'd be like, okay, this is great, and they would use it. And then they would encounter a slightly different tax issue and they would come in, and we wouldn't have a model for that problem. And they go, oh, that's too bad. I was really hoping BlueJ would have something on this and they don't. And then they might try it again, and they might have another disappointing experience. And then they're like, okay, this thing just doesn't cover enough of what I want it to. And then they forget about it, they never log in again. And so the usage was really hard to sustain. There were some power users who knew exactly what BlueJ did. They would come in all the time with reasonable frequency and get a ton of value out of the platform. But it wasn't enough of a consistent experience for all the users to really get this thing to take off. So yeah, we got it up to over $5 million in ARR over several years. It was enough to keep us encouraged, but we knew what the problem was. Couldn't answer every tax research problem that people had and so when it was actually prior to chat-GPT coming out. It was earlier that fall, it was like September and I remember the DaVinci 3 version of GPT-3 was released. And I was in the OpenAI playground playing around with it. I'm like, oh, this is surprisingly good. If you go back into the LinkedIn archives, you can probably find some of the posts that I was making at that time going, this is really interesting. Check out what DaVinci 3 is able to do. That triggered this thought, okay, maybe there's something with these new large language models, that are coming out that we can harness in order to get to this seemingly magical outcome. This magical user experience, which is if we can answer any tax research question somebody has, that would be unbelievably good. It was the original vision that we had for the company was like, let's build this thing that's tantamount to magic. That could take all of the difficult, challenging work of tax research and really automate it, and so. 

Pablo Srugo (00:19:35):
What was it, by the way? Technologically that in the old, like the supervised learning. Made you have to build model by model, case by case, but then with LLMs. It could just in theory handle everything?

Ben Alaire (00:19:44):
The big move was retrieval augmented generation, where you kind of run a search. So you have a master corpus of all of the relevant tax research materials. You run an intelligent search and you've got really good vector embeddings, and you're chunking all this stuff, and indexing it. In just the right way. You can find the relevant research resources and then you can synthesize them. And generate a very good synthesis of all of the relevant materials, and produce an answer. But you just could not do that reliably prior to the large language models having sufficient natural language, understanding, synthesis capabilities, the token limits. Especially early on, even with this new approach were challenging to work within. But The supervised machine learning approach was very laborious. We had to go through and do a whole bunch of processing of all of the case law and extract the metadata. Annotate it and train these models one by one and update them every time a new case would come out. We didn't need to update the models to reflect the new case law. The brilliant thing about leveraging retrieval augmented generation is we can just curate the data and if we're confident about the currency of the tax database, and the content in the database. And we're confident about our retrieval algorithms, our prompt engineering, the whole user interface, the user experience can be dramatically simplified and made really slick. And so that's one of the lessons that we learned from the packaging of ChatGPT was, oh, a system where you come in and people can just type whatever tax research question that they want and we are going to take on the burden of assembling all the right materials, constructing an answer, and giving them like a full takedown of the question. And provide all of the underlying authoritative legal sources for that answer. That's amazing, if we can do that, if we can deliver that.

Pablo Srugo (00:21:36):
Why can't I, if I use the latest thinking model of ChatGPT and I ask it a tax question. Why is your answer going to be so much better than theirs? What are you, what are you doing that they're not? 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.

Ben Alaire (00:21:55):
There are a number of things going on there, but increasingly you can ask the general models tax research questions and you'll often get a pretty good answer. Because it's looking now at web sources and it's able to look at a bunch of stuff. But there are a few big problems with the general models. One is they're not specifically designed with providing expert level tax research answers and they're relying on web sources. And web sources are going to be quite variable in the quality of what you find there and a lot of it will be out of date. And so if you're relying on guidance or web documents from 2022 and 2023 and 2025. Sometimes everything works out perfectly and the law hasn't really changed. And a horizontal kind of approach like a Perplexity or a ChatGPT or a Gemini will be able to cobble together an answer. And it'll be good, it'll be accurate. More commonly, there will be some of the materials that those systems find even on the open web are out of date, anachronistic, the law has changed, there's been some statutory amendment or there's some new case law or, some law firm wrote some guidance and it's now been left up on there. It's quite stale and there've been new rulings by the Canada Revenue Agency or the IRS. And that's not taken into account. And it's going to produce an answer that's missing, some really important subsequent developments. And these models don't know the difference. They will just produce a confident sounding answer and you're not able necessarily to just click into those sources and see the full text. If it's on the open web, you can but that doesn't solve the underlying problem. Which is you don't have all of the authoritative documents many of which are behind paywalls informing the answer. And so in tax law, you've got these really terrific broad content collections, and traditional publishers have assembled these tax research content collections over decades. And so the big publishers are publishers like Thomson Rueters and CCH, and LexisNexis, and Bloomberg, and they got these really great comprehensive collections of content. And they keep those content collections safely guarded behind paywalls, and so it's not on the open web. And so those horizontal tools are unable to access all of that authoritative content. So that's on informing the answers. We have copies of all the authoritative content necessary to produce really great answers. And we make it easy for our users to look at those authoritative sources and validate, okay, this is where this came from, this is where this came from, and we're focused exclusively on tax research. Which means that the quality is really a cut above what you're going to get from those horizontal solutions. And our users are very picky, they're discerning users, they're tax professionals who are not content with using, you know, a ChatGPT.

Pablo Srugo (00:24:51):
I mean, they have to be right. Yeah, they can't get stuff that's like, sixty percent right. They're looking for kind of certainty, yeah.

Ben Alaire (00:24:56):
Right and time is really valuable to them. They do not want to waste time with an inferior product. When there's something superior, purpose-built that they can access, that's going to give them the best answer. Their clients are paying them a lot of money for the time that they're spending doing this research and if they can accelerate that research with tools that cut through that task like a hot knife through butter. They want that hot knife through butter experience and that's why they're turning to BlueJ.

Pablo Srugo (00:25:21):
So walking back to that storyline, you see this just before ChatGPT. Kind of this new model, you see it could lead to that vision. How do you approach it? How all in do you go on that? Because you have an existing product with existing revenue. So how do you balance that?

Ben Alaire (00:25:34):
Yeah, it took some courage and some conviction. The courage and conviction was, the conviction was doing what we're doing is not going to scale properly. We're just not going to get that breakout product market fit that we need in order for all of this to really work out. Partly it was that, it was partly necessity. Like we're going to make this work. We have to get to some solution that can answer any tax research question people want to ask. The other one was, well, do we just forego all of this revenue? Because your point is a good one. We don't just want to tell our existing customers who are paying us good money for access and the models worked. They were providing value to the customers, but here's kind of how we navigated it. It was OK, we're going to put all of our existing tax research tools into kind of maintenance mode and we'll keep servicing the software. We'll keep updating it, but we're not going to invest in new feature development. We're not going to invest in building new models. People are going to get what's in there and we're going to take six months. I told the company, we're going to take the first six months of 2023 and we're going to focus all of our new development efforts in building this thing that can answer any tax research question in U.S. federal income tax law. Why U.S. federal income tax law? Well, the market's the biggest on earth. There are a lot of materials. We had the library necessary to build that, because of our existing relationship with TaxNotes. Which is a non-profit publisher in the U.S. and so we had access to all of the necessary content, and it was going to be a really great market opportunity.

Pablo Srugo (00:26:59):
And at this point, you had seen, like, ChatGPT had come out. And so you'd seen. You saw, the consumer adoption of that product and just how incredible it could be.

Ben Alaire (00:27:06):
Totally and I'm like, OK, there has to be this kind of version of ChatGPT, but specifically for tax research. We were in the right place, at the right time, with the right opportunity. We had a team of about fifty people. We had a decent balance sheet at the time and it was go time. We had the data science team, we had the data all prepped, we were ready to go and so we made that brave decision to make this pivot. And say we're going to use this new technology, we're going to put all of our chips on large language models and see if we can get them to do what we set out to do originally back in 2015, when we started BlueJ. And Pablo, I have to say, by the end of June. We had something that could answer U.S. federal income tax questions, but it was a little bit gnarly. It was a little bit janky. So half the time, there would be some issue with the answer. It wasn't quite right. Maybe it hallucinated something. It took nintey seconds to produce a response. It would go through the retrieval and then the generation. And it was just a single shot interaction. So if you as a user typed in your tax research question, you waited 90 seconds and you got the answer. And you weren't quite happy with it. You couldn't challenge it with a subsequent question. It was like you had to start over. So you like cut and paste the prompt, and change it a little bit, and then try again. It was OK, we had about a plus 20 NPS score. Which is like merely OK. It's the bare threshold that you kind of have to get to, to have. I think, a saleable SaaS product. But we had a long list of things we wanted to do to try to improve the performance of the system. So we ended 2023 at just over $2 million in ARR. So it was a very successful launch.

Pablo Srugo (00:28:43):
On the new product you're saying?

Ben Alaire (00:28:45):
Just on the new product. So it's like, we got to a couple million pretty fast with a pretty janky product. 2024, lots more innovation, lots more iteration. There's, of course, improvement in the underlying foundational models as well over this time, right? So we're using GPT 3.5 out of the gate and then GPT 4. We brought it on board and that started to lead to improved performance, and subsequent model developments. 4.0, we are improving retrieval, we're improving our prompting, we made it fully conversational and so by the end of 2024, The NPS was around 70. It was fully conversational. You didn't have to wait nearly as long for an answer. It was like fifteen seconds rather than ninety seconds and we grew from like roughly $2 million in ARR to just shy of $9 million in ARR by the end of 2024. It's like, okay, now this is really working. This is really getting exciting. We were cashflow positive from operations in 2024, which is pretty cool as a startup and so we headed into 2025, very optimistic. Fast forward to now, we're three quarters of the way through 2025 now. The NPS on our product is now in the mid-80s. It was trailing thirty days, it was 84, and people love it. We've gone from fewer than a hundred firms subscribing at the end of 2023 to 1,200-ish at the end of 2024. Now we're closing in on 3,400 firms who are signed up to BlueJ, and every day it's like we're adding about another ten firms. New logos every single day, and it's wild. So now we're in this hyper growth stage.

Pablo Srugo (00:30:18):
What was your growth in this last kind of year?

Ben Alaire (00:30:20):
Well, yeah, now we're in the mid 20s, millions of ARR.

Pablo Srugo (00:30:23):
Crazy.

Ben Alaire (00:30:24):
So we've already basically tripled the business from where we ended, at the end of 2024 and we still have like Q4. Which is typically traditionally the strongest quarter for this business ahead of us and a really healthy pipeline.

Pablo Srugo (00:30:36) :
That's huge. One question, by the way. Because I remember this from the older days of BlueJ and generally in legal tech. A lot of these lawyers, the way that they charge is based on, you know, how much time they spend doing something, like, billable hours. What's the effect of BlueJ on that? Does it lower their billable hours? Is it just that the markets change so much they have no choice? Or is the impact not that kind of one to one?

Ben Alaire (00:30:57):
I think what's driving it is, it really is at least ten times better than doing tax research the traditional way and so, the fact that this method exists, and it's so readily available, and the time to value is instant. When people do a trial of this thing, they get in, they try a tax research question, and especially if they have just finished a different tax research task in one of the traditional platforms that they're using. They know what the answer is, so when BlueJ produces the answer in twenty seconds and they can read it. They go, I know that's the right answer and I know how long it just took me to do this. And this other thing, which could be six or eight hours, and this thing did it in twenty seconds. They're like, okay, I can't justify it to myself to just burn my life in that way. Competitors, competing firms are adopting this thing, like if this thing exists, we kind of have to have it.

Pablo Srugo (00:31:44):
By the way, you mentioned one thing there. You mentioned time to value. I was talking to another founder recently. We were talking about growth and specifically word of mouth growth. His take, which I've been thinking a lot since then. Is that the number one biggest thing you can do for word of mouth is time to value. If you can put something in somebody's hands and they can get value right away, the likelihood that they're going to tell other people about it is way higher than if the same value comes but it takes hours or days, or whatever. Is that true in your experience?

Pablo Srugo (00:32:11):
Absolutely, I mean, the conversion, the trial conversion, the word of mouth is, it's so much fun to sell BlueJ now compared to V1. Where the value was more difficult to unlock and it was less consistent. Just think about the social dynamics there, right Pablo? Because I think you put your finger on something. Which is, if I am confident that I make this recommendation to you and you can, at very low cost, go and validate that for yourself right away. It makes it socially far less risky for me to suggest this to you. Because you can validate it for yourself extremely quickly. I'm not inviting you to go spend a few days trying to figure something out and then you might resent me for having suggested it. Go try it, just go try it, it's easy. If it works, that's awesome, because then you're going to be like pleased that I shared this tip with you and if it doesn't work well. It wasn't a huge time commitment on your part to figure out that it wasn't for you anyway but. 

Pablo Srugo (00:33:05):
Yeah, that's a great way of putting it.

Ben Alaire (00:33:07):
The consistency of the value, like I'm not asking you to spend to make a huge investment in your time and the upside gets so asymmetric. The upside versus the investment to validate it.

Pablo Srugo (00:33:17):
If you think about the opposite, like if I were to recommend Photoshop to somebody else. I'd actually have to, in order to protect that social dynamic piece. Warn them about the fact that it's a great product, but like just before you try it. It's going to take a long time to get it. All these things are going to have to. So that's going to make the word of mouth so much harder and less likely to happen than if, as you said, you know that they're going to try it and they're going to love it right away. It's just going to make you look good.

Ben Alaire (00:33:39):
Yeah, 100 percent.

Pablo Srugo (00:33:40):
There's three questions we typically end on. One of them is for you personally, when did you feel like you had true product market fit?

Ben Alaire (00:33:47):
I was doing a demonstration for a group at the Canada Revenue Agency. I think it was early 2024, and we had just built the Canadian product. And I was invited by somebody at the CRA to come into their Toronto Tax Services office, and do a bit of a talk on AI, and tax. And I did a demo. You know, I ran through a couple examples of questions, but then I turned it over to the audience and said, I put them on the spot and said. Does anyone have any questions, any tax research questions that they have? And a guy put up his hand and he said, I have one for you. He was sitting right at the front at one of these roundtables. There was probably a group of like 150-ish people at the CRS, a pretty big group and he said, this is a tricky problem. It took us two weeks to do this internally. I don't have my hopes up, but can you try this? And he described the question, and I push enter. I'm like, okay, let's see what BlueJ comes up with. And BlueJ started bringing in the answer. This guy stood up and he walked up to the screen and he was reading it line by line. That's the answer we came up with. It was a two-week research task and we had a bunch of experts. And I remember driving back to my office at the law school. Afterwards, I turned up the radio and I was just pretty pumped. Because I was like, okay, that was hugely successful. That was an ecologically valid test of BlueJ and we just hit it out of the park. It was unbelievably successful and it had all the sources. And he's like, those are the right sources. And he couldn't believe that it did it. And in those demos, the audience always has like some reservations. It's like, oh, maybe this is a cooked up example. And maybe this is being presented to really over promise what this product can actually do. Which is why I think it's always so powerful to say, well, do you have any research problems that you were working on recently? So I have no idea what you might have been working on. So to demonstrate it live to a very discerning group, that was really hugely validating.

Pablo Srugo(00:35:46):
You know, that moment you're in the car and you're just you're vibing to the music. And you feel like, hey, you really have it. These days, I'm talking to a lot of founders that frankly get a little lucky. Because they have that moment like a year after launching, right? There's just so much stuff happening. In your case, it took like eight years and you actually mentioned this earlier. You had, when the ChatGPT stuff happened right before, like that early 2023. You were in the right time, the right place, you had all the ingredients and finally, there was this inflection point in the market that you could take advantage of. But in those eight years, was there a time when or seven or eight years, whatever it was. When you questioned the whole operation, when you thought maybe it's just not going to work?

Ben Alaire (00:36:22):
It's a good question. I would tell Code this all the time, like every time I talk to Code.

Pablo Srugo (00:36:27):
Code, who by the way, is the other partner at my firm.

Ben Alaire (00:36:29):
I'm not talking to computer code, I'm talking to code the human. But I'd say, code, I'm going to make this work or I'm going to die trying. Because I had such conviction in the fact that AI could solve this problem. It was just a matter of time. Which also, that attitude of being all in and I'm going to do it or I'll die trying. It was hugely, I think, in some ways, the right attitude to set us up for seizing the moment, seizing the opportunity when it came up. it's also a little bit nuts, right? Because you can also waste your entire life having conviction about some moment that never arrives. I wouldn't necessarily recommend that strategy, but I couldn't avoid it. I just felt so viscerally that this was going to work. I would not recommend that attitude to others, but I couldn't help but have that attitude and it was just in me and so, of course, you wonder. When is this going to come? I'm certain it is going to come and when it does come, I sure as heck want it to be me who's going to seize the opportunity when it does come. And then, Pablo, in this case, I kind of feel extraordinarily fortunate. It came within time, and we didn't run out of money, and we were able to seize the opportunity. And now, of course, we're in this hyper-growth phase. Which is hugely validating, and it gets me excited about what we're going to be able to build for global tax professionals over the course of the next several years. I think it's going to totally revolutionize global tax and so I'm super excited. But we were really, really early and it was a long eight year stretch between starting out and actually getting there.

Pablo Srugo (00:38:04):
Looking back, it's like, yeah. Eight years and then things happen. But when you're living eight years, it can feel like a very long time. What would be kind of your top piece of advice for an early stage founder that is in that kind of finding product market fit phase of the business?

Ben Alaire (00:38:19):
Every context is so different. Maybe the biggest thing is be ruthlessly honest with yourself about whether you have product market fit. I think having kind of muddled through for so many years. I think it was always like we had customers, we had many, hundreds of firms signed up to the version one of the platform. So it's hard to say that we didn't have product market, but we didn't have like really good product market. It wasn't this thing where it's like, OK, this is really working and this is really solid. It always kind of felt like kind of had a partial vacuum, a partial seal and it was like. Wasn't really tight and working in the way that you would really want it to work. And it was that openness to realizing we weren't quite there. We were close to there, we had a good vision, we were attacking a worthwhile problem. Which is why we got as far as we got, realizing that there's, there's a difference when you really hit on to a very strong solution and you have to keep going for it until you get it. And then even then you can't be complacent, like now I'm still your NPS is in the mid eighties and it's like okay let's keep going. How do we make it even better like, what are the next landmarks that we have to reach in terms of the product in order to really solve tax as a domain and so we're still going. But it's being really honest about it because you're probably, as a founder, you're the easiest one to deceive. You can deceive yourself about this and you can have happy ears and listen to the folks who are trying to provide encouragement. And telling you, yeah, this is really great. I could totally see myself using this. The real test is, are they using it aggressively and every day, and thrills. And telling everybody else that they know about this thing that that's happening, and they're paying real money for it. And that's when you know that you've got that product market fit.

Pablo Srugo (00:40:13):
Perfect. Well, Ben, we'll stop it there. Thanks so much for jumping on the show, man. It's been awesome.

Ben Alaire (00:40:17):
My pleasure. Thanks for having me.

Pablo Srugo (00:40:19):
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.