How Bending Spoons Built a $20 Billion Compounder | Luca Ferrari

Can relentless optimization create as much value as breakthrough invention? Bending Spoons CEO Luca Ferrari argues that radical innovation can produce extraordinary outcomes, but systematic improvement can be more dependable, especially when paired with a model designed to compound over decades. Speaking with a16z General Partner Anish Acharya at Wave by Vento, Luca explains how Bending Spoons acquires technology companies, integrates them onto a shared platform, improves their economics, and reinvests the resulting cash into new acquisitions. They discuss the tradeoffs between growth and profitability, how to use debt without putting the company at risk, and why long-term thinking can create an advantage. They also debate whether AI really threatens SaaS, why moats like brand, proprietary data, network effects, and switching costs still matter, and how Bending Spoons is already using AI across coding, customer support, and data analysis.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Erik Torenberg?

  Bending Spoons has built a very different kind of technology company. Buy great products, improve them, generate more cash, and use that cash to do it again at a larger scale. At Wave by Vento, A16Z general partner Anish Charia sits down with Bending Spoons cofounder and CEO Luca Ferrari to unpack the thinking behind that model. They discuss when companies should bet on radical invention versus relentless optimization, why Luca prioritizes long-term free cash flow over short-term valuation, and how Bending Spoons approaches acquisitions, debt, and risk. They also get into whether AI really makes established software easier to disrupt, what

• 0:41Erik Torenberg?

  Luca thinks many SaaS modes remain durable, and how AI is already doing the vast majority of coding and data analysis inside Bending Spoons.

• 0:53Anish Acharya?

  All right. Luca, thank you for saying yes to this. Welcome.

• 0:56Speaker 02

  Thank you for having me. My pleasure.

• 0:59Anish Acharya?

  Amazing. Okay, so just to contextualize for everyone, I'm sure we're all familiar with Bending Spoons. It's an absolute monster company. I think my favorite thing about the Bending Spoon story is how cool it sounds to work there. It feels like it's got the kind of intellectual diversity of a McKinsey with the technology sort of lean-in of Google. So congratulations on building a phenomenal company.

• 1:20Speaker 02

  Well, thank you. Those are flattering comparisons.

• 1:23Anish Acharya?

  Just for the few that may not be familiar, do you want to give the 30-second overview of what it is exactly that you guys do? And then let's get into some advanced topics.

• 1:31Speaker 02

  Yeah, it's pretty simple on the face of it. We buy technology companies where we see an opportunity to make them a lot more successful, and then we integrate them into our platform, same foundational technology, same core team, and try to transform them so they're a lot more cash-generative with a, say, 10-year view, so a long horizon. And if we do that right, we take the extra earnings and lever them up and buy more bigger companies, rinse and repeat.

• 2:03Anish Acharya?

  Incredible. And do you want to note a couple of the companies that you guys have acquired?

• 2:06Speaker 02

  Yeah, so Airtable, AOL, Vimeo, Eventbrite, Kamut, WeTransfer. More than 50 acquisitions at this point, but this would be some of the more, you know, the better-known ones.

• 2:20Anish Acharya?

  Extraordinary. So you've had a really interesting sort of insider view as to how companies get built, how they plateau, and then how they start growing again. Marc Andreessen has quipped before that there's no such thing as equities. Every company is either a bond, which means it generates a lot of sort of cash for a long period of time, or it's an option, you know? It's a sort of out-of-the-money thing that may or may not happen. Do you agree with that, Luca? Do you think companies can re-accelerate once they plateaued? What do you think?

• 2:48Speaker 02

  I think Mark likes to make statements that are a bit clickbaity, and I mean it in a good way. Actually, that's a bit simplistic, but directionally true. I mean, there have been examples of companies that have dramatically changed their profiles or re-accelerated. For example, Apple with the iPhone, clearly there was a watershed moment, and that was invention, and I don't think anybody expected that. And if you run that history 10 other times, maybe eight times it doesn't happen, you know? Someone else does it earlier, or they never succeed.

• 3:29Speaker 02

  Or Nvidia recently with being at the forefront of AI. There are different worlds where Nvidia keeps being a, basically, a video game hardware, graphics hardware company, which is a great business, but obviously not something that would be worth trillions of dollars. And that was, yes, invention, but a lot of, let's say, technological tailwinds.

• 3:55Speaker 02

  So it can happen, but it's rare. So I agree broadly speaking with Mark with that statement.

• 4:00Anish Acharya?

  Do you think that investors, you know, sort of overvalue growth at the expense of things like profits, or, you know, overvalue subscribers at the sort of expense of the total size of the business?

• 4:13Speaker 02

  It depends. So I think there are investors whose understanding of the math is not necessarily 10 out of 10, and so they do miscalculate the value of some of these things. And not just the math, but the causality between different, let's say, KPIs in the long run. But a lot of investors are actually very smart, and what they're doing is they are including as a variable in the equation also what everybody else will think about the business. And so if they were to assume to never sell a share, never issue equity, but just leave off the cash flows from that business, I think they'll take different decisions than

• 4:55Speaker 02

  those they try to push management to take if they're assuming that most of the value creation will come from a sale or further rounds. So there is an element of psychologist speculation, which is very irrational. And if you account for that, then they're not necessarily irrational in their weighing of the different factors. You could argue that the fact that broadly the market thinks that way is irrational, but sometimes there are equilibria in any complex system, including the market, that you get to for historical reasons, and they are inefficient globally, universally, but

• 5:35Speaker 02

  moving out of those particular local optima is actually expensive and unlikely to happen. So I think you're probably worse off with overall the world thinking this way, but now that we're here, it may be rational for a lot of investors to play by those rules.

• 5:52Anish Acharya?

  And how about founders? You know, if you're the founder of a single product company, you're the CEO, how much do you think you should sort of adhere to this sort of distorted view of where value accrues versus focus on simple profit maximization, even if the investor doesn't feel like they get their full option on a potential sale?

• 6:12Speaker 02

  It's just a problem to solve, like all problems. You want to understand what your objective is, what the boundary conditions are, and then, once those are clear, you creatively list all the options and pick the best one. So if your goal is to, say, make as much money as possible, then it may be smart to, again, accept or play along with those expectations. As long as you know what you're trying to accomplish and why you're doing it, then you're fine.

• 6:47Speaker 02

  But there are alternatives, and they may not necessarily lead to the greatest financial outcome, or sometimes they may, it depends on the specifics. So you could choose to run a business without necessarily relying on issuing equity or selling your stake and trying to make it, in the long run, a great cash-generative enterprise that will ultimately create value for you through, for example, dividends. Or alternatively, you may work to maximize free cash flows in the long run in a way that does not get you the biggest valuation today, but will in the very long run. I think you could argue this is a little bit what we do with Bending Spoons.

• 7:31Speaker 02

  There have been multiple instances where we knew that we could do things differently to get a better valuation in the moment, do things that would please investors more, buy a certain type of asset, manage that asset in a different way. But we knew, at least we were very confident, and I think so far history has proven us right, that if we continue to optimize for the rate of compounding of free cash flows in the long run, those would be much higher than they would have been following the other path. And then even if the multiple were to turn out to be lower, you know, ultimately the multiple is just one factor in a multiplication. If the other factor is 10 times bigger, then, you know, rest assured, the overall result is going to be bigger too.

• 8:15Speaker 02

  So, again, just trying to understand what you're trying to accomplish, understand the boundary conditions, and then work backward from those.

• 8:23Anish Acharya?

  Yeah, it's interesting how much alpha there tends to be in being long-dated in your thinking, you know?

• 8:28Speaker 02

  Oh, yeah. I mean, TikTok shows you how we're not exactly, as a species, inclined to be patient and attentive and long-term oriented. So if you can do that, that already sets you apart, I think.

• 8:46Anish Acharya?

  So I've got a few other questions for you, maybe on behalf of our CEO founders out there. You know, at any incremental moment, you can sort of bet on invention or optimization. When you guys acquire something like Amiro or an Evernote, how do you decide? Do you have a rule of thumb? How do you tie-break, et cetera?

• 9:07Speaker 02

  Yeah, so I think that the...

• 9:12Speaker 02

  The type of qualities and the mindset you need to max out innovation, or at least radical invention, because I think you can be very innovative as you optimize, but I think like more radical step change innovation versus optimization incremental innovation are quite different. I would say that any company that's all in on one at the expense of the other is probably not doing as well as it can. Take, I think, Apple, for example, historically has tried to be more on the big inventions of the things, but not to the detriment of optimization excessively. I mean, famously, you know, Tim Cook was great at making sure operations were efficient.

• 9:55Speaker 02

  So you want to know, but because the qualities that you need and the mindset and the culture you need are different, I think you need to choose your camp, like are we more like the radical innovators or more the incremental innovators? And I think both are very fine, just know, pick one side and stick to it. More radical innovation will lead to higher peaks faster potentially. If you come up again with iPhone and you can deliver on it, it could transform your company close to overnight. Or, you know, Google search back in the day. Still, to this day, Google is vastly Google search. It's unbelievable. I mean, we're talking 30 years later and it's still 80%, that's still what it is.

• 10:39Speaker 02

  And they've invested tens of billions of dollars in all sorts of stuff, acquisitions, and it's still mostly search.

• 10:44Anish Acharya?

  That's right.

• 10:45Speaker 02

  So that's how powerful radical innovation is when it works, but it's highly, it's not dependable. It's very difficult to, like, the median case is super awful, basically lose everything, but occasionally you're right and you win massively. I think incremental innovation is more dependable, more measurable. And if applied with the right strategy, in the right context, like I think Ben's Pools can compound over a very long period of time. So if you, for example, run a one, a single product SaaS, I think incrementally optimizing, ultimately, there's only so much you can do. But if you do it in the context of Ben's Pools, where you get 100 in EBITDA, you turn it into 200,

• 11:24Speaker 02

  and then you lever up three times, and then you buy more, that compounds exponentially into potentially 100x over time, right? So whether you choose radical innovation as your primary pursuit or incremental innovation depends on the broader strategy you're implementing, I believe.

• 11:40Anish Acharya?

  It's interesting because we lionize radical innovation, but to your point about Google, it's almost as if, once you've invented Google, it's so hard to do that. It's rational to just optimize it forever, actually.

• 11:52Speaker 02

  Oh, yeah, you could make, I think Google is one of the best, overall, one of the most impressive companies to ever have been created. So, you know, we start from a place of incredible admiration. You could make a reasonable argument that overall shareholders would be even better off had they, you know, had the company focused fully on search and optimized search to the max. Because obviously, Google is bigger and better because it's done more, but you have to compare that with what the different shareholders would have been able to accomplish by redeploying the excess capital into other pursuits.

• 12:32Anish Acharya?

  Yes.

• 12:33Speaker 02

  And I don't know that you can make, I did the math a couple of years ago, I think it was at least a debate. We took the two would have been better.

• 12:42Anish Acharya?

  Yeah, really interesting, though we got the transformer paper out of Google, which is the foundation of everything that's happening now, so I could definitely take the other side. As you think about managing your product portfolio, how much does customer feedback and customer love play into your thinking?

• 12:58Speaker 02

  I mean, it's critical ultimately, you know, you build products for customers and that's, you know, it's customers who pay you and give you a business. So listening to what customers want is paramount. Ironically, I think that Ben's Pools does so more than most companies, while our reputation often it's more like we're more ruthless in optimizing and sometimes do things that piss off some customers. A lot of companies in my experience actually don't build things primarily because they think customers really need them, but because they think they are interesting, exciting, they're more passion driven. We try to be 100% focused on creating value for users and customers

• 13:41Speaker 02

  and then monetizing as a separate, let's say, pursuit, obviously. I think you're never going to be wrong if you spend time with customers and really understand what they're trying to accomplish and try to give them that. It doesn't mean that you take the how from them, because sometimes customers haven't put a lot of thought into it or they don't have the skills to tell you what the ideal user experience would be like. But their goals, what excites them, what they need, they tend to know better than you do if you ask and listen carefully, I think.

• 14:13Anish Acharya?

  Yeah, really interesting. I agree, a lot of things tend to get built. They get, the story gets told as if they're built for customers, but in truth, founders build things that are intellectually compelling to them. Has your taste in problems changed over the years or taste in opportunities? The type of company you want to acquire, your intuition for it?

• 14:32Speaker 02

  Maybe, I mean, I suppose so. We all learn as we go through life and do things and reflect on what we fail at and what we succeed at. Yeah, I guess not in any extremely concrete way that I can think of right now. I tend to get excited about problems in general and finding creative solutions to problems. I'm not as motivated by what, let's say, broader society may perceive to be cool.

• 15:11Speaker 02

  So my taste in companies, if that's what you mean, has not really ever been, oh, this is a sexy brand or... But if the problem is complex and solving it would create a lot of value, then I get excited. Even if the company itself, like AOL, feels old and legacy, I actually find the AOL acquisition incredibly exciting, one of the most exciting, frankly, for us.

• 15:37Anish Acharya?

  Amazing. Well, let's talk about the portfolio. You know, you had mentioned acquiring 50 companies, is that right?

• 15:42Speaker 02

  Yeah, 50 acquisitions approximately, maybe 60 at this point.

• 15:45Anish Acharya?

  Incredible. And obviously, the portfolio lets you do things that any single product CEO can't do. You know, you talked on the David Senna podcast about being able to take these expected value bets where the payoff is outsized, but the sort of downside is you may lose your job as the CEO of a single product company. Do you think that CEOs of single product companies irrationally overweight the probability of things going wrong? I mean, if you were to go back and do your single product startup again, would you change the way that you approached your investments?

• 16:15Speaker 02

  So, I guess I did co-found a single product startup and we failed. So clearly, we risked, we took a lot of risk. I think the short answer is no. I think there are probably executive teams that substantially miscalculate risks. But I think primarily, the reason why they don't take some risks isn't because they, you know, they got them wrong by 20%, but because they hate the median case and they should. Look, it's like if you're a VC, and you may be better positioned than me to elaborate on this,

• 17:00Speaker 02

  but you may want, ultimately, you invest in, say, 50 companies. What you care about is the overall returns from the portfolio and you may want each of these companies to go 100 miles per hour against the wall. And you know that the four or five who get to go through the door, which is very narrow, will win 100x and the others will crash and burn. Whereas if they had taken it more slowly, they would have lost the opportunity of winning, but they probably would have had better chances of being somewhat sustainable, decent businesses. For you as a VC, the more cautious approach is bad because your expected outcome gets worse and your median outcome doesn't actually get much better.

• 17:45Anish Acharya?

  Partially better, right.

• 17:46Speaker 02

  But for an individual company, while their expected outcome would be better if they went 100 miles per hour, their median outcome would be massively worse. It could go from, I end up with something to I end up with nothing, right? And so it's not irrational for some of these teams, whether it's entrepreneurs, professional managers, to try to protect the downside, against the downside a little bit, because ultimately, that's, you know, they only have one life and one career.

• 18:11Anish Acharya?

  I mean, that's what I appreciate so much about what you guys do, Lukas, is that for, you know, extraordinary companies that have been innovative and have achieved scale, you know, but perhaps haven't gone all the way to the stratosphere, they have an opportunity to keep building and creating value with Benning Spoons.

• 18:26Speaker 02

  Yeah, I think we are a very important, and perhaps missing up to now, element of a healthy innovation ecosystem. Proof being that, you know, every time we have made an offer to someone to buy their company, either they chose not to sell or they sold it to us, not to someone else. So when you have such overwhelming, called a success rate, it means you're really finding great product market fit in a way, like you're really solving a problem that people have.

• 18:57Anish Acharya?

  Yeah, yeah, absolutely. So you've talked a bunch about sort of bending spoons in terms of the focus on clear strategy and extreme focus on talent. It feels like credit and capital markets has been a key part of what you guys have achieved as well. Can you talk about some of the lessons from leveraging debt effectively? Because some companies have gotten in trouble and yet you guys have benefited and been so strategic in how you use debt and debt markets?

• 19:22Speaker 02

  Sure. So that helps you basically do more sooner, faster. And assuming your expected returns from those investments are higher than the cost of debt, it's always smart to take that with the with one big caveat, which is assuming the risk of defaulting on your debt remains negligible. Because obviously, if it becomes significant, then it's not a no brainer any longer that taking the extra risk makes sense. Now, as an intellectually honest disclaimer, I think if you take, you know, companies that at some point went bust, but hadn't had issues up to a certain point, and you ask them, they would all come across

• 20:06Speaker 02

  as very bullish about their usage of debt. So we have been doing fine and have not even come remotely close to having issues with it. But it's still important to remain cautious because it's not the people who went bust thought they would as they raised the debt. I mean, I guess in some cases, there were some perverse incentives to still raise the debt. But in most cases, they thought they were fine. And then things turned and they had issues. So I think it's important to be paranoid in general when taking on any significant amount of debt. We have stringent limits we self impose, trying to keep leverage ratio well below 3X. It's currently about 2.5, 2.4. Historically, it's been closer

• 20:50Speaker 02

  to 2 than to 3. Always been in a position to be able to pay back 100% of our debt before maturity, even under severely adverse scenarios. We try to refinance most of our debt every year or two to keep maturity five plus years out. We hedge 100% of our debt against interest rate fluctuations. So we know exactly how much interest we will pay in the future. We're highly diversified. We have a very high level of profitability and high cash conversion. So there are all sorts of things that mitigate the risks associated

• 21:33Speaker 02

  with leveraging up. And my assessment right now is that our risk of defaulting on our debt is close to zero. It's extremely, extremely low. And so we're better off having that debt than than not. But again, be cautious. Don't overdo it, particularly if you're not sophisticated and you don't fully understand the cash flow consequences of it. Start small, develop knowledge, good modeling, good processes and checks and balances, and then maybe take a little bit more. But again, any project that could be huge would probably require a decade plus to get there. And rule number one to win and be that huge success is not having failed in the meantime. And so I think often it's better to take two, three years

• 22:17Speaker 02

  longer, but take no or very limited risk than speeding up and then crashing and burning.

• 22:24Anish Acharya?

  I mean, I think these are great rules of thumb, you know, perhaps encouraging a bit more ambition in product and the size of the swings and a bit more caution in credit and debt. We've seen many situations where founders tend to underestimate the probability of things going sideways with debt.

• 22:40Speaker 02

  Particularly if you rely on one segment, one market, one product, one revenue stream, because boundary conditions can change quite abruptly. You're still not fully insulated if you have a diversified portfolio, because diversification is a complex concept and it depends on whether you're diversified or not, it depends on against what particular phenomenon and you could, you know, your different supposedly diversified assets could be highly correlated when it comes to the response to that particular phenomenon. But so again, handle with care. It's tricky. But if done properly, it's a massive boon. I mean, there's plenty of companies would use that consistently in history, you know,

• 23:23Speaker 02

  Broadcom, TransDigm, and have done incredibly well for many decades. And they wouldn't have done nearly as well had they not used that. So.

• 23:32Anish Acharya?

  Yeah, interesting. My theory is that everything is correlated, but that's a subject for a different talk.

• 23:38Speaker 02

  Well, I tend to agree. I mean, but not necessarily to the same extent.

• 23:42Anish Acharya?

  That's right. That's right. Okay, let's quickly talk about durability in moats is a favorite topic in Silicon Valley right now. You know, a year ago was SaaS apocalypse around Opus 4546. All those names traded down 40% in backup. So I don't know what folks accomplished with the SaaS apocalypse. It seems like it was a little overblown, but it has brought about this discussion of where durability lies in a world where you can trivially create software with AI. How do you think about it?

• 24:08Speaker 02

  I never fully. So I think the valuations for SaaS historically were, for the most part, absurdly high, but not because AI was not a risk, because they didn't make any sense from, at least from the perspective of someone, going back to what we were talking about earlier, someone who values assets based on reasonable expectations of free cash flows in the long run. In most cases, I think you could only justify those valuations based on the speculation that you'd be able to sell them to someone for, say, the same multiple down the line. But in the meantime, you have grown 50% or something, right? Because they were based often on assuming that the current 12% organic growth would last for 15 years. And it's like, sure, but you know, the world changes so quickly. That's

• 24:53Speaker 02

  insane, right? Organic growth almost never goes up, by the way. I mean, very rarely. Again, going back to Marx, it typically goes down. And so those assumptions made no sense. Now, I think AI gave people an excuse to go back to their models and say, maybe I don't fully believe in those assumptions. Let me be more cautious. And the reset valuations were, I think, broadly speaking, reasonable. I mean, you could always argue plus minus 20%, but they were more in the range of reasonable, in my view. Now, I never really bought too much into the risk of now writing software is cheaper, therefore these companies are screwed. I do buy in the idea of it's going to be much cheaper, therefore they'll

• 25:35Speaker 02

  suffer on the margin, but now they're screwed. And the reason is, I think, writing software has been visibly cheap for a long time. And most of the companies that matter, you know, scaled companies worth at least, I don't know, a billion dollars, the cost of replicating their product technically

• 25:50Anish Acharya?

  Yes.

• 25:51Speaker 02

  was negligible relative to the value of being a winner or gaining significant market share in that market even five or 10 years ago. Like, yeah, maybe you needed 20 good engineers, but replicating sales force is not rocket science. We own WeTransfer. If you had 10, you know, five, five great new graduates who want to spend three months in a garage, they could probably build a product that's 80% as good. You know, and in fact, guess what? 100 teams have done that. There's plenty of close, right? So mostly software or digital products do not succeed because it's very expensive to replicate

• 26:31Speaker 02

  them. They're now SpaceX. They mostly succeed because of other factors. First mover advantage. Sometimes you start, you build a brand simply because people know you and nothing else. And then there is a lot of habits people don't, you know, actual switching costs, but just also the psychological switching costs. People don't want to bother switching if it means, you know, our company making a billion in revenue, you're saving $3,000 for that SaaS. Who wants to even bother taking the time and risk? And it's rational not to do so. Two, it could be network effects. We own Eventbrite. People don't go to Eventbrite because, you know, nobody among all Homo sapiens has managed to replicate a ticketing

• 27:13Speaker 02

  platform. But it's because on Eventbrite, you have 90 million people who go monthly look for for tickets. And so you sell more, basically, you know, right? And so AI can replicate that a million times, but they don't have the audience. And people go there because everybody posts events there, right? Or something like Commut, you have proprietary data. With Commut, because of the history with all these users, we know that that particular path is muddy and, you know, it rained and there is a dog barking. And we have photos of those corners and trees and whatnot and turns that are not available. They're not available on satellite. They're nowhere to be found. And so, again, it's not, it's actually, Commut is quite sophisticated, actually, the routing, but still, you know,

• 27:55Speaker 02

  smart people can build that over time, but they don't have the data. Or scale advantages, you know, Airtable, our most recent acquisition, most customers pay Airtable less than $20,000 a year. Mostly they're large enterprises. Even if you get to a point where you can ask your agent, AI agent, you press a button and it replicates the code for Airtable pretty much perfectly, which we're very far from, but we'll get there, I think, you know, soon enough, probably in the next few years.

• 28:25Speaker 02

  Supposedly, you'll still want to have someone at the company who's responsible to making sure that that code seems legit, that if something breaks, goes to the agent and asks the agent to fix it and whatnot. I think there's a far future where AIs run our lives and companies, nobody works any longer, which is interesting, but you know, it's not, I think most people will agree it's not in the next five years.

• 28:45Anish Acharya?

  That's right.

• 28:46Speaker 02

  As long as you have someone being responsible for things, even just a 0.2 FT of a software engineer would still cost you more than paying air table for the same thing. So sure, it's an interesting philosophical point, but ultimately you're not going to do it rationally. It's dumb. So there are all sorts of reasons why I think most, not all, but most digital products are reasonably well protected. Again, they will suffer on the margin, new customers, you know, maybe people search for products through their agents and agents are, have a different set of criteria they follow to choose. And so before this, they, people tended to choose product A, now it's more product B, that will certainly hurt product A, but, but not to the, to the point that they melt away overnight, you know, because of this.

• 29:29Anish Acharya?

  And those are natural market evolutions, right? Like web to mobile and preferences change, and maybe it's agents instead of humans that are searching. But no, our data shows that the enterprise spends eight to 10% of total spend on software. So as you think about where you want to point your AI and your token budget, it's probably not in replicating payroll. That just seems like a high downside, low upside proposition.

• 29:51Speaker 02

  I so agree, yeah.

• 29:53Anish Acharya?

  Excellent. Hey, so you mentioned proprietary data, which I think is actually under discussed these days with everything happening in models. You know, there's a quip that many companies are sort of trading for less than the value of their tokens, which is the value that they would have if they were simply acquired for training data by models. Do you think that's true? Do you think proprietary data has changed?

• 30:13Speaker 02

  So I, I haven't thought enough about this and I don't know enough about it, but I will say that claim strikes me as potentially a little bit misleading. Meaning the, the value of data within new pools declines pretty rapidly. It's, it's all, it's like if today I write a new blog post on the internet, nobody's going to give me money for that. It's private and I'll go to open it. I'm about to, you know, I may publish it, I may not. Do you want it? Do you give me it on top? You know, the first blog post they ever got access to or could ever access, they would probably have paid a thousand dollars for. But, but now zero, because incrementally it doesn't add anything to, to the corpus of knowledge that you can train AI models. So yes, maybe there's now a need for plenty

• 30:56Speaker 02

  of data on how people do project, project management or whatnot. The first companies selling that maybe make a bunch of money, but I think over time that trends to zero as the, you know, the quality of incremental training, you know, returns deteriorate for incremental training. But I'm, I'm in a position to determine whether, you know, companies could sell data for more of their enterprise value. I'm not, I'm not certain about that.

• 31:23Anish Acharya?

  Interesting. Let's double click on the kind of operating system you use within Bending Spoons to help improve and optimize all of the companies. Can you talk about some of the key components of it? And which of the components do you think would have been most useful to you as a single product founder versus are very, you know, specific to what you do with the portfolio?

• 31:42Speaker 02

  So, so there's pretty much everything that you need to run a digital business. We're built in-house with the exception of highly, I think, universal tools like, you know, Gmail or Slack, these sort of things for you, or Google Drive. We're happy to, to use those that others build. We don't have highly specific or personalized needs. But everything else from data processing, the A B testing platform, all the marketing automations, user lifetime prediction, credentials management, recruiting tools. It, the list goes on and on. Depending on how you count, it's over 50 different in-house tools. They, we build them ourselves.

• 32:26Speaker 02

  And, and the reason why this has been valuable for us is, well, the, the obvious way is we, we save on vendor costs, quite a lot actually, but, but it's not something I'd say is a game changer.

• 32:37Speaker 02

  And the two game changers are, one, we make them super fit for purpose and very specific to our needs and they're, and these tools are natively integrated with one another. So they, they, the sum of the parts is much greater than what you could possibly accomplish by buying these different tools from different vendors. Cause each is better for what we exactly need to do and they also work in a more harmonious way together. Also because we can, we, there are this big monolithic foundation, we can swap out the foundation at an acquired business quite rapidly. It's actually much faster to change the entire thing in one, or more or less the entire thing in one shot than removing one piece

• 33:19Speaker 02

  and then you place the, the alternative and you need to plug it to everything else. Then remove another piece, plug it to everything else. Actually swapping the entire thing out tends to be, to be faster. Now, of course, this is not particularly relevant to a single product company because they, they don't need to buy new companies. So that's, that's not what they need. But they could probably benefit from a bunch of these tools. A lot of the value is unlocked by having the right culture and the right talent. Cause like every tool, ultimately the user makes a ton of difference. And so depending on what company we're talking about, they could be able to take advantage of it or not. But some are more, let's say autonomous and they don't require a lot of effort and quite powerful. For example, we're built something called Alt, A L T, like alternative Spooner that's

• 34:03Speaker 02

  a, it's basically a, an agent that lives in Slack and you talk to an as, automatically has access to all the same tools and, and assets as you do. It's the same GitHub repos, same folders and drive and whatnot. And, and you can task it to do things for you. Anything you could do in theory, you could do. And increasingly it can do and does. So you can tell it, for example, to run a data analysis for you and it will go, fetch the data. If it doesn't have access, it will escalate a request to our credentials management team and potentially our IT security team if, if relevant. Get, get the data, process it, show your graph and commentary. You don't have to ask anybody else. You get it in five

• 34:47Speaker 02

  minutes, save many, many hours and certainly a lot of delay in the exchange of information. I experienced an interesting situation where general manager of, of Evernote for us in, in a channel where we post feedback on the product, she wrote to her agent, I've seen this bug, can you please go to Moro, which is our customer support tool to check whether other customers have reported it, then go to the code base, find the root cause, code a fix and ping Marco, which is our engineering lead on, on that product to review the fix and fix it. And, and it got done and that would, I mean, it, it was a hundred times better than what was possible before. So that's maybe a tool that I think could be colossal game changer for a lot of companies.

• 35:32Anish Acharya?

  That sounds really powerful. As you think about the way models get deployed within the company, do you think there are any job functions where the model will be doing 100% of the work in the near future? Or do you imagine it being a kind of lever for individual productivity as you just described?

• 35:46Speaker 02

  A 100%. So a customer support, for example, we have evolved our, all of our customer support managers from being more of like agents answering, you know, responding to, to tickets. Now we are, for that, we pretty much 100% use either AI or for the residual portion of the tickets, external vendors. And our customer support managers are more and more similar to say a product manager trying to architect the ideal response process, response rules and user experience. And this is just the beginning, you know, we're investing heavily in this to make it much better. We'll include at some point, support via phone

• 36:30Speaker 02

  calls, via WhatsApp. So that, that's an obvious strength for L L M's because it's mostly about digesting a body of knowledge and, and exchanging communications with people and L L M's are, have been trained on data that's quite similar to this and, and already today are pretty good at it. Another area where we, we have seen the percentage of, of tasks taken by AI increase massively is data science, data analysis. Now our data scientists are doing much more here to architectural work, building, writing software and building tooling and processes for say the budgeting process or a recurring dashboard. But they, the, the

• 37:15Speaker 02

  actual fetching data, cleaning data, plotting data type of work is done. I don't have a precise figure, but I'm pretty confident at least 90% by AI. And then coding, coding, of course. We were at 0% written by AI two years ago, where now, last time I checked at 94%.

• 37:34Anish Acharya?

  Wow.

• 37:34Speaker 02

  Some companies, you know, depending on how aggressive you want to be, you could be, if you're a good company, it could be at 70 and 99, but no good company should be, I don't know, below 70 right now, in my view. There's no point.

• 37:46Anish Acharya?

  I agree. Everyone should be converging to 100. Okay. I've got a couple more questions here, Luke, and then we'll wrap it out. So, you know, one question is, you've talked about how within Bending Spoons, every decision should be able to make with maximum rationality, but that doesn't necessarily mean always being led by data and analytics. And L L M's can also often create a sort of false precision. Is there a type of decision where you found being instincts led versus analytics led has led to better outcomes?

• 38:13Speaker 02

  So I think analytics and instincts are inputs. So data is an input. My instinct that something is off or this could be cool is an input. You want to get as many inputs as possible within reason. But then I don't think there's ever any case in life where you should be deciding with any other than, anything other than logic and rationality. Again, it doesn't mean that simply because an analysis shows that, you know, people in France are more likely to buy this, then we do that and we do it in France. It's just, again, one input in a broader discussion. But I don't think there's ever a good reason to say, we'll do this because I feel like it, you know, unless we're talking leisure, which is all about feeling a certain way.

• 38:57Anish Acharya?

  Right.

• 38:58Speaker 02

  Even if you are the most, you want to be as creative and imaginative as possible, but again, those are just inputs. Those were produced ideas or options that you wouldn't otherwise have considered, but you don't pick those simply because they, you got them in the spur of a moment and you're super excited about them. You want to ponder them and determine whether that excitement is justified. And if so, yeah, go for it. Absolutely.

• 39:20Anish Acharya?

  That makes sense. So maybe one question on behalf of founders who feel like they're stuck. You know, the definition of hell is never winning and never quitting. How should founders know when it's time to kind of move on versus, and of course, this is a difficult question, versus, hey, maybe you're three months away from the next increment?

• 39:35Speaker 02

  Yeah, I got it. I wish I had a good answer. You know, I'm probably awful at that because my startup before bending spoons ever tail, we were, worked at it for three years and, you know how we failed? So it clearly wasn't working. We were trying to use AI way too early, clearly, to write diaries for people automatically. Clearly wasn't working. People weren't using it, certainly weren't paying for it. And, but we kept pushing, we raised a little extra money to keep going and pivoted and pivoted and pivoted.

• 40:10Speaker 02

  And then at some point, we were left with maybe something like, maybe $250,000, $300,000. But much of that was meant to be through a, like a tax return from the Danish government. At the time, we were based in Denmark. I don't remember exactly the bureaucratic aspects of it, but something like that. And then we got a message from the government that refund either wasn't due ultimately or was due 12 months later, something that would basically, that put us out of business. And so they, you could argue that red tape made the decision for us. So I'm not very good at that. I feel I can't advise.

• 40:51Anish Acharya?

  Bureaucracy created bending spoons.

• 40:53Speaker 02

  Yeah.

• 40:54Anish Acharya?

  You heard it here first. Okay, final question. Look, for the next technology company that comes from Italy or even from the region, what do you think they should learn from bending spoons and what do you think that, you know, they should think about for themselves?

• 41:06Speaker 02

  They should definitely think with their own heads. 100% there's no, I've never seen a great entrepreneur, a great inventor, a great engineer, a great artist, a great anything who doesn't ultimately take responsibility for their own decisions and is intellectually independent. But some things that I feel are almost universally useful, at least in entrepreneurship and business or ambition. I feel, look, ultimately, it's very subjective. Personally, I think that you, if you can only do so many things and almost anything is more interesting, is a greater likelihood of ultimately being rewarding if you aim high.

• 41:47Speaker 02

  So just aim for, you know, now, aim for the moon, like that's, that's lost all its flavor because of the crypto guys, but aim high, aim very high. Two, again, try to, try to be rational, be as creative, imaginative, passionate as you can be, but ultimately try to make decisions using, using logic that will, that will get you much further. And by the way, some of the biggest inventors in, in history do that pretty well. I think Jobs did that well, certainly Bill Gates did that well, Elon Musk clearly, you know, does that well, Bezos. They're all guys with big ideas, but ultimately they don't pick them because they say, well, that sounds so cool, let's do it.

• 42:29Speaker 02

  They, they look at the facts and all the, and at the first principles and ultimately counter decision of those. So I'd say ambition, rationality, and then going lower by one level, but really do not underestimate how important it is to have a talent dense team and how impossibly hard it is to improve talent density if you started with very low talent density. For a bunch of reasons, we probably don't have time to discuss, but most, most of those should be fairly intuitive.

• 43:02Anish Acharya?

  Look, you built one of the most unique and successful companies in the world. Congratulations and thank you for spending time with us today.

• 43:08Speaker 02

  My pleasure.

• 43:09Erik Torenberg?

  Thanks for listening to this episode of the A16Z Podcast. If you liked this episode, be sure to like, comment, subscribe, leave us a rating or review, and share it with your friends and family. For more episodes, go to YouTube, Apple Podcasts, and Spotify. Follow us on X at A16Z and subscribe to our Substack at A16Z.substack.com. Thanks again for listening and I'll see you in the next episode. This information is for educational purposes only and is not a recommendation to buy, hold, or sell any investment or financial product. This podcast has been produced by a third party and may include paid promotional advertisements,

• 43:52Erik Torenberg?

  other company references, and individuals unaffiliated with A16Z. Such advertisements, companies, and individuals are not endorsed by AH Capital Management LLC, A16Z, or any of its affiliates. Information is from sources deemed reliable on the date of publication, but A16Z does not guarantee its accuracy.