Join discussions on whether the traditional unicorn model still holds as emerging firms push beyond billion-dollar marks toward “super-unicorn” status.
At Davos 2026, leaders and investors argued that a $1B valuation is less a finish line than a noisy signal whose meaning depends on fundamentals. Rob Heyvaert said a “plain clean billion” conveys belief in future cash flows and credibility, but warned against excess: “Don’t give a founder too much money,” especially when valuations are propped up by “structure.” Christina Kosmowski viewed unicorn status as enterprise reassurance, yet insisted proof arrives at renewal: “You don’t really know until the renewal cycle comes up.” Matthew Fitzpatrick quantified the shift: unicorns grew from 40 (2013) to 1,600 today, driven by AI-era scaling and extreme power laws that “pull forward” valuations and shorten time-to-unicorn. Panelists converged on durability metrics—retention and gross margin—over growth theatrics. Fitzpatrick cautioned that revenue acquisition is “overrated” when compute-heavy models create “perpetually negative margin” businesses; sustainable winners pair “happy customers” with margins. Bootstrapped founder Nacho De Marco contrasted customer discipline with fundraising “dopamine,” arguing customers punish mistakes faster than investors. On AI economics, debate centered on “forward deployed engineering”: necessary for enterprise adoption but potentially margin-dilutive. Leadership lessons: ignore valuation “noise,” prioritize focus, and recruit elite talent—often easier at lower valuations. As De Marco put it, “Leaders are paid to describe the future.”
Good morning and welcome. It is Friday morning, the last day of a very consequential annual meeting of the World Economic Forum in 2026. I'm. I'm the founder of Exponential View. We help investors make sense of the AI economy. And I'm absolutely thrilled to be moderating this session on how high can unicorns fly. Another way to think about this is do you know what's cool, $1 billion? Well, that is what used to be cool many years ago, and perhaps it's still cool and important today in the innovation economy for startups and for investors looking at disruptive and breakthrough companies. Now, to explore this question, I'm really delighted that we have this expert panel with us, starting with Rob Hayward, who is the founder and managing partner of Motiv Partners in the US. Rob has built two unicorns himself. He invests in unicorns today and I'm eager to hear his views. Next to him is Christina Kozlowski, the chief executive officer of Logic Monitor in the US, and she has a great deal of experience of seeing companies go from a unicorn state to a unicorn state next to her. Matthew Fitzpatrick, the CEO of Invisible Technologies, working in the AI observability domain. And finally, Nacho DeMarco, the CEO of Bizdev, a unicorn who didn't seek external capital for experts, for remarkable perspectives. I think it's important to think about where this word unicorn came from. It emerged out of Silicon Valley a little more than 12, 13 years ago, as a way of identifying companies that had broken through that had, in some sense found a path to scale and potentially a path to profitability. But for many of us in the sector, sometimes it felt like it was a status that entrepreneurs were chasing a badge of honor that might not necessarily lead to sustainable businesses. We can think of many, many cases, such cases that we, I'm sure will discuss. The question is, what does it mean today? Is it a useful sentiment? Does it tell us anything about the sustainability of those businesses? So let's perhaps start with you, Rob. What does that $1 billion valuation actually signal today?
I mean, if they do their homework well, and when we say $1 billion valuation, have to be careful that we don't potentially use structured instruments to get there. So but if it's a plain clean billion dollar valuation, it means you get a business plan. It means you have a team that has a track record potentially. But more importantly, there's a belief in the future. And I hate to say it, but even a billion, 10 billion, 5 billion, it's all about future DCF. And the investors discipline is on the one hand. On the other hand, in my humble opinion, what's more important is we need to continue to have abundance of capital going into the entrepreneurial circles, and it needs to be at a scale and at a respect where they can really disrupt themselves. And I think a high valuation, lots of access to capital potentially gives you a chance to disrupt an industry versus saying, oh, you're not relevant. Who are you? Can you even survive? No, no, I've got these four backers. My last valuation was 3.5 billion. So I think that's very powerful.
The credibility that may come with that. Christina, what does it signal to you? The billion dollar valuation?
Yeah, absolutely. I try not to get too caught up in the valuations, and especially keeping your employees focused on what matters and what you can control. But with that said, your customers, we have very big, large enterprise customers. And so when they invest in logic Monitor, they want to ensure that you're going to be around for a long time. And so it gives you that credibility with your customers and that validation.
Matthew, what's your experience been?
So let me put some numbers behind this. In 2013 there were 40 unicorns in the world. Today there are 1600. So 40 x increase over a decade. Now the question is what is driving that? And I'll give three different reasons. I think for it. One, scaling laws are different in the in the AI era. Like I think if you look at the fact 50% of the total capital is going to AI, the recent class of Y Combinator is double the revenue base of any class before it. And so the rate at which revenue can scale here is different. The second issue I would say is power law dynamics. So a couple stats on that. If you look at the top decile of seed funded startups, are seven X more likely to become a unicorn than the bottom decile. And then of the remainder as you get to the growth stage, 30% of total capital is going to five companies, and 70% of capital is going to 100 companies. So you have an extreme power law dynamic paired with scaling laws. And I think what that is leading to, and this is the long answer to your question, is that people are pulling up the are moving forward, the valuation. So basically the time to unicorns. Ten years ago, the average time unicorn was three and a half years. Today sorry it's eight years today. It's three and a half years. And so effectively what's happening in the DCF math is because of that power dynamic and because of those scaling laws. People are saying, and I guess you could say in some ways, unicorns mean less as a result of that. But people are pulling forward valuations because of capital scarcity and power law dynamics. And they are there is a large amount of capital trying to funnel into a small number of companies. And I do think what that has led to is the definition of a unicorn has changed in this era.
There's a really interesting sub thought in there about how risk has has shifted. If you're getting $1 billion valuation from, from from the get go, which we'll return to in a second, I'd just love to hear from your your perspective on what that billion dollar signals.
Well, first of all, I'd like to have your voice, Matt, and your confidence and data. But I have one thing that I believe billion dollars mean, which is customer validation. They trust you. You've created something that actually creates value for customers. To me, when we reach that valuation years ago, the first thing that came to mind is it's just a number, really, it's just a number. But at the end of the day, it basically means that you're creating value. You don't get to that number without having created real value for your customers. And I think that clear validation is really what in a way brings me joy. Other than that, it's just a number.
You got to your billion dollar valuation essentially through bootstrapping. So when you say $1 billion was validation by by customers, and if that's always the case, a company that raises money at a $10 billion valuation on two sheets of a Google doc, has got validation from someone. But in that case, are the customers, the venture capitalists who've met that validation valuation?
I think founders many times confuse fundraising with with progress. And and, you know, fundraising gives you dopamine, right. But real progress gives you customers. Right. You may you may start thinking that just because you've raised money, you're actually creating money. And there is a correlation. But it's not always the case. Right. You've seen you know, you're in the investment world. I am my you know, my side job is being an investor with about 80 portfolio companies today. And reality is that many of those companies will not get anywhere. Right. But if you're building something your customers want, you're solving a problem. You're creating value for them. Chances are you're going to be around and then raising money may be a misleading, metric, perhaps. And, you know, you're going to get a lot of dopamine, but that doesn't mean you're actually creating creating value.
I love that I've written it down. It has to go out on social fundraising, gives you dopamine. Customers give you value.
Do it with Spanglish, please. Okay.
Oh, well, a little bit of Spanglish. So, Rob, I want to come in to you.
So what I what I would say is, over the years.
I've learned this. Don't give a founder too much money. And there's one to have a valuation. But I've seen, by the way, those who remember the.com crisis, there were tons of money. Somebody with an idea could raise $200 million, and he spent it on the wrong thing. So the combination of disciplined capital have abundance of it. But then be very disciplined in how you deploy it is extremely important, right?
Extremely important. So that is and I think that that's one of the things that we might observe as being tough in today's AI environment. I mean, I've seen any number, probably more than a dozen term sheets come my way where it's $1 billion valuation and there's not a customer sometimes on a product yet to show. Matthew, I was so interested in what you said about that dynamic playing out. So it feels like the risk, more risk is being taken earlier stage, not merely because the valuation is higher, but to to Rob's point, the amount of money that's coming in can lead to a psychological indiscipline on the part of the founding team.
So I think it depends on on the business model within AI. And what I'd say is, to Rob's point on capital, I think there's one area of AI, particularly foundation model training. So LM builders, where the bet people are making is that 90% of the capital used there is going to go to compute cost, right. So it's the way to think about if they're betting $1 billion up front is they're betting on a team and the use of kind of all the available tooling and then spend on compute to build something. So it's a different bet than, say, an enterprise business, which we're in, applied AI enterprise company. And in our case we don't spend a ton on compute. So there's still though I do think in the I r a need to raise capital, I think if you are in a, in an environment where there is this much competition and this much activity, the challenge of the growth versus harvest stage is you can you can raise less and bootstrap. And by the way, I mean many amazing businesses have been built that way. But there's also the idea that you want to scale faster and build moats. And so I think it's almost two different dynamics. I think in the LM model building world, you need the capital to just spend on compute. In the enterprise. You're doing it to build products and moats and in a competitive landscape. And so I do think that's the kind of the flip side of the current landscape is if everyone else is raising money, you do have to think about the degree to which you raise comparables.
Yeah.
The only thing you have to be careful because when you say a billion, most of the terms that go out of the billion have structure. That's right. So what does that actually mean at the end of the day? Because the money's going to come out first from the one that came in. So I think you have to kind of also be careful. I think it's not all about to your point valuation.
And Christina, I'm eager to your, career involved a couple of companies that grew really fast, famously for their customer revenue and their customer retention. Right. Salesforce and then and then slack. So what are the lessons that you've you've learned there that perhaps when you look at the market today, you think these are these have been unlearned?
Yeah. I mean I think, you know, I say you don't really know until the renewal cycle comes up. And so I look at a lot of these, you know, early unicorns. I'm like, have they had a renewal? Because that's when the true value gets tested. And so I think that's really important. And you've got to stay focused on do you have product market fit. And I agree with you that yes, you want to grow fast and acquire customers. But again, if you're just spending so much, it's really hard to tell if you have this true product market fit. And that's what you know. I know I always keeps me up at night at logicmonitor. I'm like, do we have market fit? Is the macro conditions just, you know, fueling. So like, what is it what is actually the value proposition here. And so I think that's what.
I just ask you to unpack that for me slightly. There. So you said that, have you really got true product market fit? Well, in the old days, if you had a growing IRR, annualized recurring revenue, you had, a good retention. Your customer acquisition costs were not so bad. That was a sign of a sustainable business. What are the metrics? Do those are you suggesting that those metrics perhaps aren't enough right now in this frothy market?
Yeah, I mean, I think they still are. I think we haven't seen this customer acquisition being cheap. Right, like or efficient. I think people are doing whatever it takes to go get customer acquisition. And then again, we haven't seen the renewal or retention rates. And then in this market especially anything that's AI. People are throwing money at it because they're being asked to experiment without actually understanding what they're what they're trying to achieve.
Matthew, please. Yeah.
No. Look, I actually think it's a great point. I think that in the current market, acquisition of revenue is overrated. Like, I do think it is easier now to acquire a lot of revenue and then not be able to deliver it. I think that is one of the challenges. But some people have called revenue. I actually think the biggest marker is if you think of a valuation long term, a DCF is determined by how fast you scale the revenue and the persistence of that revenue. And so in my mind, actually, the true I completely agree with you. The true markers of success are retention, customer happiness, and retention. If you have happy customers that are sustained for a long time, that is how you build a real business. And I do think gross margin is the other one. I think there are a ton of companies right now that are negative margin and going to be perpetually negative margin, because they're just using compute as a way to deliver what they have. And so I do think those are the two things the sustainable businesses will have real gross margin and retention.
Let's peel those two apart. We'll come to gross margins in a second. I want to come to the sustainable bit of the question. If you're looking at metrics that are based on DCF or or retention, that is a metric that develops very, very slowly over time. And yet decisions are having to be made more quickly because of the, the competition. So there's a there's a paradox or there's a compromise that needs to be made there. You're having to deploy capital at this moment. So how do you.
Yeah it's funny. It is all DCF, but you make it up because it's a piece of paper. You just you have no clue what's going to happen. So I think we in a lot of it's more about the people and the markets and the ability to imagine what could be possible if those people attack that market with the same passion that an entrepreneur does and have this unique angle, and then it's still a lot of intuition, in my humble opinion. And then, of course, you build from there. Pivoting is also a big thing. I've seen so many companies pivot, and I think that's so important. So it's all it's all about that ability to be to be super agile in my opinion, because DCF is is discounted for a reason.
And nature. You've built your your business bootstrapped. You do have AI in your name though. So I think that was very prescient, of you when you.
Not planned, by the way.
It's intuition. So so just think about what you see in the portfolios that you're investing in personally. And how do you bridge the fact that you've built a business in a, what we might say, a traditional way, through customer revenue.
So, so interestingly, I feel like there's a there's a significant difference in the way you build a business when you're fully bootstrapped versus when you raised, money, in a way, fundamentally, you see a difference where, well, first of all, you're freaking out about making payroll every day. That's that's obviously one one clear difference, but but when you think about it, I think many times VC backed leaders are. Lean towards moving really fast, even if the unit economics just don't work. You don't have a business that works. Right. And but you have.
To explain unit economics. If people are unfamiliar.
Like we talked about gross margins, we talked about customer acquisition costs, right? Those type of things when you have no money, which is our case. When bears started 17 years ago, we had really no money, zero zero. And so in a way, if it doesn't work for the customer and the customer is not seeing value, I'm not going to pursue it because I can't. Right. And fundamentally, when you have men in your money, in your bank, you're in a situation where you may lean towards, oh, I'm going to bring revenue by spending more money than the revenue that will come in. Right. And that's something that it's very hard not to do because the pressure to do that many times, they're obviously very smart investors who may not give you that pressure, put that pressure on you. But but oftentimes founders fall into that trap of just having a business that doesn't really work. Instead of pivoting and finding a way to bring real value so that they can actually have economics that work well for their business long term, that's something that it's very easy to avoid when you have no money.
Yeah, it reminds me of something one of my mentors told me. He said, I've seen many more businesses startups die of indigestion than starvation. Because the starvation puts discipline in you. I said we'd come back to this question about about margins, which I think is so important in this world of very, very fast growing AI companies. And I was reflecting on the fact that about a year and a half ago, the measure was AI companies are getting to $10 million of annual recurring revenue three times faster than similar SaaS companies a few years ago. But the charts now show getting to $200 million in annualized recurring revenue, not ten, because nobody is interested seemingly in that ten. But whether that economics is sustainable is a different question. And Matthew, you raised this point about gross margins. So how did AI economics margins, compute costs, data expenses, talent expenses differ from traditional software or software as a service businesses? Do you want to jump in? I'll come to Christine.
Yeah, sure. I think call out a couple of differences. So the theory of software is that you build it once, you never have to touch it, and it should run at 95% margins forever. Now the reality of software is that's never how it's actually been. If you look at all the public knows as well, if you look at all the software multiples or companies, they all have tons of services, they all have tons of forked code. And one interesting fact during this era is SaaS. SaaS multiples have, in the public markets, gone from 20 to 10 X in the last three years because they've tried to become profitable. Software actually struggles to get profitable. But the theory of it is it should be 95% margin for error. AI is not that. I think there are a couple of big differences AI one more consumption based unit based revenue rather than pure platform fees. So our revenue is a mix of platform fees and consumption, but you're going to have more outcomes that are tied to usage or outcomes. Revenue is not just going to be tied to platform fees in the air. It's just not a feasible it's not a logical way to to structure a lot of deals. Two, you do have more ongoing costs tied to compute. So, there's huge variation in that. There are some companies, like many of the coding apps, that have very large amounts of compute for their revenue. Most of the say the applied enterprise companies are using more inference based. So compute is kind of single digit percentages of their revenue. But that is an ongoing cost. And then AI does have more configuration and for deployed engineering around it too. And we're very open about this. We do all our own configuration and deployed engineering because you can't just sell a box and say, this is my AI box, take it. You have to configure it and make it customized to the individual company. And so and customer. And I think that's been the big change in this era is people are realizing this is no longer about I invest 100 into a box. And then I, in theory scale that box forever, which really didn't actually ever work as a model. And it is more about customization.
I just want to help you. You help me clarify a couple of points. Come to Christina and then. But you said, just explain the forward deployed engineer why that is what that is, why that's relevant and why people have got so excited about it in the last year or so.
Yeah. So I think as many people in the room know, traditional software, the way you would do it is you'd have you'd have a box, let's say you're an insuretech and that that insurance piece of insurance technology do one thing, and what you'd have is like an Accenture or a cognizant would come in and they would do all the data linkages to map to the schema and stand up your box. You can't really do that in AI because usually the workflow, if you take something, I'll take insurance like claims processing, very specific to the institution, you're generating something that's not just numbers. It may be like a 20 page document or a 30 page document. It requires training, reinforcement learning, validation, and a lot of workflow customization. And so what what Ford deployed engineering involved in Palantir was kind of the company that that really started this many years ago is embedding engineers alongside the customer in the early days of configuration. So when we do a deployment, we actually start by putting 3 to 4 engineers, usually 2 to 4 engineers on site at the client premise. So physically in person, sitting next to them embedded and stand up the the application there and then it becomes software over time. But it does start with configuration.
I mean, back in the late 90s when I was in enterprise software, we used to have that, but you'd go with the CD-ROM to the clients and to to install.
And the one thing I would say is different about like the evolution of that is you can't rely on a services vendor to do that work for you. I do think that's one of the differences. If you if you say my business model relies on Accenture or Cognizant to scale, you're going to be waiting a long time. And so I think the reason AI has had much more deployed engineering is you need to control your own customer relationship to get it stood up.
But there's one more clarification I need from you. Just a brief one, because you said something that I think surprised me and it may surprise other people. You said there's lots of forked code in there. What do you mean by forked code in a SaaS business, and why would we be surprised?
Well, so every time, like the theory again would be and Rob, you probably know more about this than anyone. You build a piece of software that is supposed to work one way, but then you talk to a customer and they're like, I need this feature, I need this feature, I need this additional feature, I need this workflow to look this way. And so what people used to do is write all these really complex stored procedures. And what that's led to in SaaS over the last 20 years is 50 different versions of every different deployment that when you that when you try and do like each incremental version annually, you can't do you can't really push that through cleanly because there are different pieces of software. And so I think in many ways the big lie of software over the last 20 years is it's not a single monolithic tool. It ends up becoming 50 very custom tools which require a lot of support, customization and a really, really hard to update every year, which is why you have so much legacy. 70% of the software in the world is over 20 years old, and most of it is terrible for that.
Sounds like a great market opportunity that 70%. Just quickly to Christina, if I may, and I'll come to you Rob.
So I sort of disagree a little bit because the forward deployed engineers, you as a software company are putting that's why the margins are bad. And are they ever going to get better. I don't I don't know that they are because you're training your customer, that you're taking on all the risk versus in the software. You're saying if you want to customize and go be a special snowflake, that's fine, but you go do it with an external consultant. And so the risk then shifts to the customer. And I think there's a lot of value in saying are forcing the customer to think, what are the best practices here versus allowing them to create this special snowflake software. Exactly.
For I'm going to just bring Rob in here as well on this, this point. So, so that we don't dive too deeply into software business models just on hearing that that discussion there does feel like there's a different underlying economics in AI based companies with all the attention is what do you see as an investor.
For sure.
And if anything, I think traditional SaaS companies, I think are in trouble, that model in my long term. And why? Because if you think what AI can do for you is it's the whole flow of complex workflows and it's potentially reinventing your whole enterprise. And so and that's a that's a whole set of other economics where it might be outcome driven. So we don't know. We're experimenting with companies trying to do better pricing around that. But it's not it's not actually there yet. But I think the economics could become far more interesting for the winners. And I think the traditional multibillion dollar SaaS companies are struggling, and I don't know if they can transform as easily as they say they have everybody locked in because it's all legacy. But what happens if you say, okay, no, no, I'm thinking a decade. I'm not thinking three days or three years, I'm thinking a decade. And what what does a bank look like in a decade from now? And that is you can't make it up. But but but it's interesting. I think the economics are going to become super interesting and there's going to be less pressure on margin, but more pressure on actual value creation, on the workflow and the execution of of a capability. We're going to decompose whole industries and decompose them again, and they're going to be winners and they're going to be unicorns that that help create that. As a result of that.
Fantastic. What we'll do, I want to just change tack a little bit to talk about the leadership dynamic of this, of this question. And how do your roles as leaders change as the companies grow? And is this, this mythical valuation? Maybe it's the unicorn. Maybe it's the I think we've heard of Decacorn now. And is there a is there a giga corn emerging somewhere? Yeah, there there must be. But and Christina, I'm really interested in whether there is a transition point, that a company goes through as it, as it grows, perhaps as it passes certain revenue milestones or valuation milestones and what that means for founder in terms of trying to run that business and keeping the team on side.
Yeah, I mean, it's it's significant, right. Because you can no longer go in and do everything. So you've got to figure out what's most important at this stage that we're at that I need to spend my time on. And so I at the beginning of every year say, all right, this year is strategic partnerships. So I've talked a lot about at this moment in time, the ecosystem is so critical. And so that's something that I uniquely can do. But delivering all the customer outcomes and all the processes I need to allow my leaders to kind of go there. And so being really clear around, what are you going to focus on? Where do you need these repeatable processes and where are you going to, you know, push your push your leaders on it.
Is is there a moment where this becomes noise, the external attention that a company startup might have if it's seen as being high flying Salesforce, slack, your your new company?
I think you know, even when I was at those companies that Mark and Stuart were very clear like that is noise. Like that's not what we're focused on. And I think we the macro economic valuation multiples and all that that changes and comes and goes. So you have to understand what is most important to our business. Are we staying focused on delivering the value to our customers. And are we making our bets in a controlled manner? I will say, as you get larger, you can say the when you make a bet to unwind, a bet becomes harder. So when you're smaller, you can pivot easier. But when you're, you know, we're 400 million in revenue right now. And if I'm making a bet, it could take me 12 months to unwind the skill set of the engineers and things that I'm building. And so I have to be much more controlled about what we're doing there.
And congratulations on $400 million in revenue. That's a nice number. Nacho, you will have faced a different pressures as your company. What your company grew. Let's just start with this one. Which is. At what point did did this did your business stop being about, did you start to have confidence that the business was actually working rather than a sort of expensive quest you'd put yourself on?
You're not going to like this answer. I think that will never, ever happen. At least to me. I think it's a it's a bit of a mistake. Paranoia for founders is a is a great, great thing. I often say that we're we're basically paid to describe the future as leaders. We're paid to describe the future. And hopefully we're going to be right more times than wrong. But but another thing that I believe every founder or leader should do is be really, really paranoid and operate like if you're going to go bankrupt tomorrow, right? Now that said, I think there was a time, fundamentally, I would normally go to the doctor before making payroll and say, if I get a heart attack, please bring me back to life, right? One day I was like, oh, it's just a calendar reminder. We're good. So that's one thing. Call it financial success, right? That there's a moment in time where you basically don't worry about making payroll. I think that moment you're good as a bootstrap company, that moment you're good. It's life changing. You come from scarcity to, okay, how do we become a decacorn or giga corn? And and then perhaps there's a more subjective metric, which in our case was seeing that all the software we were building for our clients was really out there, right. Hey, check your phone. Oh, that's a cool app. I've done that. Oh, look at your computer. Oh, we did that right. That's perhaps more of a understanding that you're having an impact, a real impact in the market. Consumers are getting better products because of you. That was perhaps the the moment I felt like, okay, we're really doing something here.
You're part of an ecosystem where you'll know many venture backed founders are you built that scale without any valuation pressure. Do you think that venture backed founders are being systematically misled when they're trying to build their companies?
There's a number of things where I think your customers will punish you for your mistakes way faster than your investors. Right. And no offense, I'm an investor myself. There's this joke. Hey, mom, how did we become so rich? By dad charging two and 20 to underperform the S&P 500, right? Right. And and I'm terrible at jokes, but. But what I'm saying is, in a way, they have multiple bets. Your clients have only one bet if you don't build exactly what they need and they don't make money out of what you help them build, they're going to punish you. They're going to punish you very, very quickly. Whereas your, you know, investors won't necessarily punish you as much. Right. And then there's, you know, we talked about this dopamine. Right. But I think there's also this pressure for, for speed that doesn't necessarily always lead you to the to the right place. Yes. Sure. Going fast is great, but doing it in a way where it's sustainable, it's even more important in my view.
I think the,
The investors attitude is also changing and the role is evolving from, spray and pray to really saying, okay, can I help you? Can I be impactful on your client acquisition talent and so on, but genuinely not as a as a tagline to get my money and then structure you all the way to. And that's why I think, by the way, venture capital is not a great capital allocation strategy. It's not if you look at if you put it into, it's too much risk and so on. So I think we're going to a world where I think the transformation I talked about in total industries, you're going to see pools of capital come in at every level, and you are going to have a choice and say, okay, I don't want that. I don't want the structure. And these are the four people I want to work with. And if you work with those four people, the chances of you being successful is much higher. Yeah.
Matthew, can I just ask you from the leadership side, unpicking these, this tension that keeps emerging between the discipline that not having capital provides you, the discipline that customers who need this one thing that works for them provides you against the, the opportunity to go quickly, the opportunity to have a strong balance sheet that gives you some strategic alternatives. But then we heard from Christina that you get to a stage where the balance sheet is not the thing that constrains your ability to, to pivot sitting within that. What are the types of decisions that you are trying to make and which are the ones do you find particularly difficult to do? Well, yeah.
So I have an interesting perspective on this because before leading invisible, I led McKinsey's internal development group for a decade. So we were building tons of different applications for many different clients, but off of opex. So off of the PNL effectively. And I do think the way you build is very different when you're doing that. And I've been through many of what you're describing, which is you are constantly thinking about within a year, how do I think about the spend within a year and how it will pay back within a year? So I think the liberating part about having capital is it gives you the ability to make long term, two, three, four year, often an AI that may be a 6 to 12 month bet because of the pace of moves, but to invest capital in things? I will say though, I think the concept of valuation to kind of what Rob is saying is a little bit overrated because it's a point in time moment. It's not an actual realization of any kind. Meaning if you build something that's worth a lot and it goes to zero, that accomplishes nothing. And so I think our North Star has not really been to think about valuation at all, really on the journey, we're thinking about building something that five, ten years from now, we want lots of recurring revenue customers that are very happy, a really good team. I spent a ton of my time thinking about recruiting, so like one, one kind of controversial take I would have is that capital allocation is less important than recruiting elite people, particularly in the I world. If you can recruit the best engineers, they will find the things to build. And I do spend capital allocation is number two. But I would take recruiting over capital allocation right now.
And think about this is the lower the valuation, the more attractive is to attract talent.
By the way.
Not the other way around.
It's actually a really good point. And so in many ways, like being obsessed with incremental milestones on the journey is the exact wrong way to do it. Rob. Exactly right. You can actually get much more talent in if you're if your valuation is lower. And so I think if you keep the North Star of where are we going to be in five years, where are we going in ten years? And your organization is thinking that way, you're fine. And you have to find investors that think like that. There are a ton of investors that will come in and say, I'll give you some big number, but like there's a there's a very famous one who makes you triple your projections every time they take you take their money. Everyone knows who that is. Well, you're making a bet. You know exactly what you're doing. You're going to have to triple all your projections and try and hyperscale, and you'll probably erode the core of the business. So our view is long term sustainable enterprise value, recurring revenue, happy customers. That is the North Star. And everything in between doesn't matter.
That is why, as Natalie said, leaders are paid to describe the future. We have time perhaps for some questions as well. I'm happy to continue discussing with the panelists, but if you if anybody would like to, in a few minutes, I'll come to to the floor and you can put your hand up and a mic will will come to you while while the the group considers whether they want to intervene. Let's just pick up on on this this particular theme. And we've talked about leadership. Now I'd like to just come back to, the AI wave of the moment, because it does feel to me that it is so significant in in everyone's decision making. And the shape of companies seems to be changing the every generation, the number of people you need to deliver revenue diminishes. General Motors needed more employees than Microsoft. Microsoft needed more employees than Facebook. Facebook needed more employees than Instagram. When it was an independent company. Instagram needed more employees than WhatsApp when it was an independent company. And that is, I guess, the theory of automation being being writ large. And in Silicon Valley there is a discussion about the one person unicorn. Is the one person unicorn plausible? Rob, why don't you.
Start? Yes.
Yeah.
Now, again, what does it mean? We just talked 20 minutes about the unicorn issue. If somebody has an amazing, amazing moat and uses all the infrastructure, but because people are investing gazillions of dollars, and if you can be a smart entrepreneur, sitting on top of that and using the right connectivity, you could have this amazing idea with four people in a garage. By the way, some of the best software is written in 2 or 3 weeks and 2 or 3 months. Not in ten years.
Yeah. So so and I'll tighten the definition by unicorn. Well let's let's assume it's not a fantasy structured valuation but it is underpinned by, you know, nine figures of revenue. For the sake of argument, Kristina, do you think that one person unicorn is is possible?
Well, I think it depends, right? I mean, I think it's not possible in enterprise software, so we won't see it there in consumer software. Yeah, it is plausible. But you know, one of the things we've been talking about this week is this concept of, is the technology you're building for people bringing joy like human, is it improving human experiences. And so I just wonder, whatever product you're delivering, if you're not having some human that's helping ensure that you're actually making another human's life more enjoyable or higher quality versus just doing something if you can, if you can get that effect.
Okay, so not in enterprise, maybe if you can make someone's life better, but just briefly because we'll go to questions straight after that.
I would say potentially in consumer. Absolutely not in enterprise. And I mean, you look at the stats right now, 5% of models in enterprise make it to production right now. So you actually have a huge adoption problem or capability problem in enterprise. And the only way to solve that is be really good sellers or people who know sector experts for deployed engineering, good product leaders. That is all like we're a very human centric company. I think humans are going to be more and more important to get enterprise AI working. So again, maybe in consumer but not enterprise, I think these teams will look actually maybe even bigger than some of the traditional early teams in SaaS.
What do you think?
I hope so, I hope so for for a simple reason. You know, we're we're a US company. We're conceived in Latin America where access to capital still still a challenge, right. And this is true for we live in a bubble. Almost everyone in our crowd, no offense. We all live in a bit of a bubble. We are. You know, this is the most relevant conference in the world. Probably the fact that we're here means that, you know, our reality is significantly better than the large majority of the population, right. And I think AI brings something that I wish we had when we started, which is the possibility of actually building something impactful, whether you get to a unicorn valuation or whether you don't, does it really matter? But you can build things that were unthinkable years ago, and one person can have a real impact without capital. That for humanity should be absolutely phenomenal, where we're going to benefit from all the brilliant people around the world that don't have the means today. Now they do. All of a sudden they have a power. They didn't exist before. So I hope so. I really.
Hope so. Wonderful. I'm going to go to the floor, Rob. We'll come back. There's a lady at the front who'd like to ask a question. If you could just briefly state your name and ask a single question, that would be wonderful. Yes.
I'm Rebecca Boudreau, CEO of Oberon Fields, and the valuation is a number, as you've discussed. But the value of a company is defined by the people who help build the company. What are your some thoughts on your the team that you build over the life cycle of creating and building your unicorn?
Yeah. Who would like to take that first?
Yeah. I mean, what's interesting is, I'm really believe and I have we have portfolio companies where the founder has 90% and we have portfolio companies where the the capital was distributed toward the shares very early on. I'm personally a very big believer of sharing broader in the broader context and be smart about it, because you have to reinvent your management team three times to get to the 400 or 5 times. But I'm a big fan of sharing, and for me, it's a red flag. If if the founder sits on his shares and actually is very prove it first, I think trust is something you give in the beginning and not at the end. And when it comes to transformation.
I'm going to ask you just briefly, because you didn't take the external capital, you probably didn't have to negotiate with a founder to say, what's our Esop percentage at this round? How have you approached this question?
So to me, you know, I'm in the people business, right. And and we hire brilliant people for, for a living. Right. So obviously I'm a little bit biased towards what does hiring the right people mean? Well, for us it means everything really. But if I wasn't running various dev and I was running any other company, I would still say that our job as leaders is to find the best people to join you. If you're not spending as a leader at the minimum ten hours a week interviewing yourself, you should reconsider your position. I think you should not be leading a company. There's nothing, nothing more important than having brilliant people. Why? Because your model will change. The world changes, right? How do you solve for that? By having brilliant, hard working people around you. And I think that's really, genuinely what matters the most for any company in the world.
Thank you, Matthew, very briefly, because I want to get another question from the other.
I think simply with both capital and people, people tend to sometimes I think founders will overemphasize dilution. So like giving away a portion of a company. What I think you should think about it as being multiplicative. So if you take capital and spend it well, you will grow. The value of the company grows in aggregate. I think people are even more extreme example of that. I go back to the idea. The number one differentiator in AI right now is recruiting. So if you can recruit great talent, you should pay that great talent well because it will grow the aggregate value of the company materially. So people who are hoarding share values and trying to keep it all for themselves will not be successful in AI. The number one thing, and we are our employee base owns about 60% of the company right now because we were we were bootstrapped for a long time. It is the most important thing to getting a really dedicated and amazing team is making sure that equity is broadly distributed.
Thank you. There's another question up at the front. Gentleman here. If you just state your name and ask one clear question.
I'm a university professor of Nankai University, China. I just want to know what's the role of technology in this context of unicorn development? You mentioned a lot about, talents, customers and so on and so forth. What is the role of technology itself?
Christina, do you want to go?
I'll jump in. My investor talks a lot about the rule, like we're moving from rule of 40 to rule of 70. And so I completely agree that we have we're talking about, you know, the external we're selling an AI product, but we have to use these products ourselves so that we can scale faster and not just even more efficiently, but go get, customers that we could never get. We've got some major, you know, multibillion dollar revenue competitors. And it was impossible a year ago to be able to go catch them. And now I'm like, we can catch them because now we can generate pipeline faster than we've ever been able to generate that. And there's limitless possibilities. And so it's very exciting. But I'm relying on my partners, my technology partners, to help me do that.
I'll come to you.
Yeah. I think if you think about scaling laws. So like when SAP was built in Germany in the 80s, they literally had to go to a factory and on a mainframe, build an initial system that took them two years. So technology was very painful back then. Now you can spin up a cloud instance, spin up an LLM, you can use the coding tools. And so your ability to build the velocity of that building is 30, 40, 50 x, what it was 20 or 30 years ago. And I do think that's why scaling laws have changed. So a lot of what you're thinking about is as a CEO is how do I allocate my capital across the different products I'm building? But the rate of particularly if you're using the coding tools, which are if you have great engineers and coding tools, you are a ten, 20, 30 x faster builder than any traditional software team. So I think that's why I emphasize recruiting so much is if you find the right people, the velocity you get from those people is infinite in the current technology environment, and that is a function of the tools available.
Now, Rob, I'm going to come to you for the last word. We have just under 35 seconds.
I'll be very brief.
I think it changes our industry completely. If you look at what my guys are working on is really we have we do judgment. We sit in a, in a long meeting called Investment Committee, and we have all these memos and we have Bain. We have everybody have McKinsey. And then we make a judgment call. I think that's changing. We are trying to see if if there is such a thing where we can create an infrastructure where the machine makes the decision for us and we've been experimenting. It's phenomenal. It's early days, but I can imagine fast forward 5 to 10 years from now, that commoditization of of insightful investing is coming. People don't like it, so you cannot put the two and 20 who said that, but I think it should stay a while. Two and 20 if you don't mind. As long.
As a bit longer. Yeah. Well thank you. Thank you very much. This has been absolutely incredible discussion and I'm so sorry that we have run out of time. I simply cannot sum up the points that have been raised today. There is so much insight in here. I'm going to encourage you all to go back and watch the recording when the forum makes it available. Leaders, we learned, are paid to describe the future. I want to thank our four leaders on the panel today who have described very, very compelling perspectives on the next coming years.
Thank you. Thank you very much. Thank you.