Luca Ferrari，Bending Spoons CEO：从4万美元起步，买下产品市场契合度，击败私募股权

We could have another presidential moment here on the All-In podcast. So if she takes it upon herself—if she calls you—just hang up. I have a phone here, just in case.

A lot of us have been following you for a while now. I first heard about you because you were in Milan, where my wife's family is from, and you had this incredibly progressive, methodical approach to growth. You did this fantastic podcast with Patrick O'Shaughnessy, which was great—I urge all of you to listen to it—and you explained the arc of Bending Spoons.

I would like you to tell people about the first few years, all the suffering and failures, the lowest point of the company, and then the beginning of the rise.

1. Crashing an AI startup, the $40,000 restart & buying product market fit

Most entrepreneurs experience a lot of pain, I think, but the biggest failures were in the previous startup. My co-founders and I launched an AI company in 2010—very early, obviously. We went bankrupt 3 years later. We had about $40,000 of capital left that we had raised from the venture capital fund, and there was nothing to salvage other than our relationship becoming stronger.

The venture capital fund gave us the money because they didn't want to go through the liquidation process, with all the legal fees and headaches. They saw how hard we worked and told us, “Keep it to yourself. We'll sell you our shares at a par value of $1, and you can go have a nice vacation.” We were clearly a little bit unhealthy, so we took the money and enthusiastically turned it into seed funding for Bending Spoons.

We started with exactly $40,000 in 2013, and we had a strategy that has remained pretty much unchanged. Obviously, over time you become smarter and improve it, but the idea was that we're not very good at finding product-market fit—or maybe luck plays a big role. Probably both are true.

We've become pretty good at engineering, design, monetization, and marketing through 3 years of hard work, and so we should be among the best in the world at those things. We should be able to buy the product-market fit from people: they get a good price, we get a good asset that we can make more valuable, and then we invest more capital in making our platform more competitive.

How much did you pay for the purchase, and how did you close the deal?

The first purchase was approximately $10,000. It was a mobile app specifically for the iPhone that you used to personalize your keyboard. A developer sold it to us. It was obviously a very amateur operation, and at that time it wasn't that difficult to make it better and more successful.

What did you buy? Did you buy disposable income? Did you buy the app? I guess you bought the one that had a small income.

It wasn't even monetized.

Good, which of course never happens for a big business. But it had users.

It had users. What we bought at that time was an app with a bunch of users and good positioning in the app stores, so it would get an influx of new users. That has generally remained unchanged over time. We continue to look for great brands, user bases, and customers that we can improve in every way and ideally make more valuable over time.

We just do it on a much larger scale today, but the basic concepts haven't changed.

2. The in-house tech stack, shrinking the teams & the 10 out of 10 standard

Did you rebuild that app? Did you take the codebase and refactor it? Help us understand technically what's going on in the organization, from this business to some of the bigger ones today. Do you do engineering, product, design, and marketing—all of the above?

Yes, we rewrote it completely, but those were very early days. Today, we are much more complex. We implement, so to speak, an operating system with more than 50 proprietary technologies.

We've created a kind of engine for managing technology businesses very efficiently and effectively. In fact, our main product is to replace the technology foundation of the businesses we buy so that we can manage them much better. The people we transfer between our different businesses also always play by the same rules and are more effective because they find the same tools.

Do these tools cover, for example, HR, finance, technical operations, and DevOps?

Almost everything. Orchestration of artificial intelligence models—checked. Recruitment tools—tested. A/B testing platforms—verified.

And then you move all the technology costs to the higher company so that you have one agreement with AWS and all the licensing scales within one organization.

Yes, this is a lever for creating quality. I would say it's a relatively small number that probably adds, I don't know, 1 to 10 percentage points to EBITDA margin. The more important aspects are the ability to increase revenue through a better product that can be monetized, sometimes marketing, reducing costs through smaller teams, meaning teams with more talent. And yes, supplier optimization is useful.

You used Elon Musk's scheme before he did. I mean, there are several stories that have been written about how you got the Vimeo workforce right and got the right people in place. No, I'm kidding. Oh, he did it? Oh, no. Good advice.

But explain how you reduce the number of people. How did you figure out that you could cut 80% of the team and it would still work? How did you understand that? Is it a coincidence that you just reach a certain threshold?

I think that's something we learned, partly because in the beginning, when we were buying smaller businesses, these people would usually sell us the assets—the product—but not the team. They wanted to move on to whatever other project they had. So we didn't really know any better. We created internal teams to continue the work, and the number of full-time employees was much smaller than originally planned.

Then, when we eventually bought businesses with established teams, we may have naively created teams to run comparable businesses that were much smaller, so we couldn't explain why you necessarily needed more people.

Partly through experimentation, we found what you might call the golden mean. Obviously, it's never perfect, but we generally want our businesses to perform at a 10-out-of-10 level. We've found that you're more likely to achieve that level of performance if you have very small teams, an extremely high bar for talent, and smart ownership.

You have a market cap of about $40 billion.

Well, right now it seems to be plus or minus—I don't really know. I haven't checked the ticker since our IPO.

Roughly around that range, which is incredible considering it started out as a $10,000 purchase. When did you go from scaling cash flow to using debt and more sophisticated financial engineering so that you could go after these bigger fish? What did your finances look like?

Historically, we started using debt in 2017, I think—either 2017 or 2018. They were very simple bank loans, TLA. Reinvesting free cash flow has always been relevant to us. We've reallocated almost 100% of our free cash flow to acquisitions since the beginning.

Since 2017, as we've scaled, we've become more robust, a little wiser, and more sophisticated. We started using TLBs, and maybe in the future there will be bond issues and other more complex instruments.

We didn't use a lot of equity. In fact, when we did the IPO, we raised only about half a billion dollars of primary equity, while our valuation was about $20 billion. Even that half a billion dollars was raised almost entirely in the previous 6 months or so.

3. Debt as an accelerant, what happens if rates rise & who else is bidding

Almost all of our track record was achieved through reinvesting free cash flow. But I think in the future, especially as a public company, tactical use of equity here and there might be a good idea.

With that equity, you borrow, I think, 5 or 6 points above LIBOR, so 10% to 12% loans, and then you buy a business like Airtable. That means you have to make $100 million in debt payments a year. If interest rates are rising—and there is a 90% chance that they will start rising—what will that do to the business? Is it slowing down a bit?

And then my second question: People have been very enthusiastic about the progress you're making.

And I think you're facing some competition from Bending Spoons right now. Maybe you could talk about whether you see more people at these auctions, and whether it's not just you and the other 2 players.

Yes, debt is a growth accelerator. We would still grow pretty quickly if we were just using free cash flow, but certainly being able to use debt is good—at a sensible level.

I would say there are 2 parts. Firstly, the risk associated with existing debt: all of our debt is currently at an average blended cost of about 9%, plus or minus, and it's fully hedged. So a rise in interest rates will not affect our cost of debt. The maturity date is in 2031, so we have the opportunity to fully repay the debt before maturity. We currently have about half the leverage.

If interest rates rise significantly, new debt will be more expensive. I think in most scenarios this would be a net positive outcome for us for a couple of reasons. Our unlevered returns have historically been quite high, consistently above 25%—again, without leverage. So whether we pay 9% or 12%, of course I would prefer to pay 9%, but that's not the case. It doesn't violate the model.

The second aspect is that typically, when interest rates go up, asset values go down. As a consistent buyer, I think we're likely to benefit more from lower valuations than from higher debt. It depends—I'm generalizing and simplifying a bit—but overall, we think we're pretty well protected and reliable when it comes to this.

When it comes to competing for acquisitions, in all the processes we've been involved in, there have been other buyers in almost all of them. I'm sure the competition will intensify—or it may intensify, who knows. It can also weaken. We see private equity has historically been down, so they've raised less capital to do the same thing, and overall we may be in a better position.

It's also important to note that it's very painful and time-consuming to replicate what we've created, because a lot of it is based on technologies that can't be built overnight. You don't even know what to actually build unless you've gone through many years of painful experimentation, trial and error, and repetition.

A significant part of the value we create comes from these approximately 800 people. We've carefully selected them over time, along with the high-performance culture and scientific approach to business that we've developed. These are things for which there is no shortcut.

I still remember how I hired first 1 person, then 2, then 4. You could probably do it in 5 years, not 13, but not in 2 months. So I think we will face competition, but I am quite optimistic about it.

I think in our industry—in the venture capital industry, and even in going public—we're looking for a founder. If a company loses its founder and his authority, it has, like Elon, the opportunity to say, “Hey, we're not going to make a Model X. We're not going to make a Model S. We're going to turn them into Optimus.” Only founders make such bold bets.

You have a slightly different philosophy here. You don't want the founders to be inside the company. You're not looking for founder authority in each of these brands, from what I've heard from you. What do you expect from your brands? Do you want cutting-edge versions of Eventbrite and Vimeo, or do you just want them to grow at a predictable rate and dump that cash flow? Let's talk about the role of the founder.

If you have a founder with that level of passion and that mentality, then 9 times out of 10 it will be a significant positive outcome. Typically, when we buy companies, these are businesses that have been around for 10 or 20 years, and in some cases even more than 20 years.

For founders, if they're still on board—and sometimes they're not—it's really a moment of, “Okay, this is a chapter I'm closing. I'm going to move on.” So the real question is: What do we gain from having this business work better with us than it did under the previous owners?

Of course, if we could have exceptional founders who would stay with us and put their soul into the business, that would be even better. But we can still succeed by being a better home for the business than if it stayed with the same ownership group and perhaps lost the founder anyway.

It's not that we don't need founders, but once companies are sold, people tend to be eager to move on.

4. Inside the deal desk: what gets acquired, why they don't build & the founder question

If you invite us into the M&A department—for example, into the room—walk us through your selection process. How do we do it? What are we looking for? Are we looking for synergies and integration with assets we've purchased previously, or are we looking solely at cash flow? How do we rank these things? Take us through the deals department.

There are qualitative criteria that we use to narrow down a long list of businesses that would be interesting targets. One of them is scale. We use a very deep integration process, and deep transformation requires a lot of operational effort, so we can't do a million of these.

By the way, the amount of time and effort required to transform a business doesn't really scale linearly with revenue. We're much better off acquiring a relatively small number of large companies than a million small ones. So we're looking for scale.

We look for predictability in earnings, and that's a big topic in itself, but we like businesses where we're pretty confident that we can predict their development for at least the next 5 or 6 years. Then we look for businesses where we can create a lot of value. It could be technology, organization, product, monetization, or marketing—ideally, most of them.

Does value include integration with these other assets that you have, or does value just mean economic value from operating activities?

Let me ask a detailed question. For example, if you own AOL, you can place ads for Vimeo, Eventbrite, or Miro on AOL. If you own Vimeo, you probably have a sales team that sells advertising on AOL that you could use. How strong is the synergistic effect?

If the synergy effect exists, and you have all these capabilities for design, creation, product management, agent orchestration, and A/B testing, why not also create organically at the same time and leverage the network effects of existing businesses?

Historically, we've created almost no value from the customer-facing synergies that you described. We've created a lot of behind-the-scenes synergies, as I said, all built on the same technology foundations, and there's this big core team of people that we're moving seamlessly between them.

Going forward—and the reason we haven't unlocked a lot of value through customer-centric synergies is because, in my opinion, the portfolio hasn't necessarily been large enough for good overlaps to materialize. But as it grows more and more—for example, Airtable and Miro are now quite attractive to many businesses—I think what you're describing could be an additional dimension of value creation.

You haven't tried it, or you tried it and it didn't work?

No, we have tried it, and it worked, but only marginally.

So maybe it helped 3%, but it wasn't the bulk of the value. The bulk of it brought 10-out-of-10 excellence in operations, product, monetization, and technology.

Then why not create organic products?

On an individual business basis, first of all, you can't do everything. I mean, alone, I can't. My colleagues and I don't think we can. Maybe we should be more ambitious about ourselves, but you have a lot to do.

There are already so many different types of products. We are launching many new products in addition to existing brands.

I understand, but these are not exactly radical innovations.

We don't do this much. We try to focus on 1 thing, to try to be the best in the world at it.

Also, at the scale we're at now—with Miro close to $4 billion in revenue—this is difficult. For example, if you look at the percentage of new startups or products being launched that could break even, it's very, very small. We would have to bring in a lot of our resources, and it's unlikely to work.

Can you just talk about what you and I were talking about—the talent drain that's happening in Silicon Valley companies as they start to stall? The talent working for the business is perhaps not the quality of talent that you've built on your core platform. How much of this is valued in that mergers and acquisitions process that Chamath mentioned?

It's hard to assess from the outside, but you can form a principled opinion. Businesses that are in the saturation phase tend to be less attractive to some of the most enterprising engineers or designers. So you can assume that the talent level will be maybe good, but perhaps not as good as Entropic.

They have a very unique type of talent, but we'll talk about that another time.

Yes, okay. We have a big advantage in attracting talent, because if you work at Bending Spoons, it might be one of the few places in the world where you can spend, say, 1 year rebuilding email infrastructure for AOL, then 7 months helping to reimagine subscriptions for Vevo, and then building a technology platform for managing payments—all with the same employer, mostly the same coworkers, and the same culture.

You get a wide technical spectrum, so there are career opportunities. You stay motivated because it's interesting and new. Very high talent density begets high talent density, so there is an element of a virtuous cycle.

Last year, we managed to get 800,000 applications.

We hired fewer than 300 people.

Everyone in Milan? Where are they?

No, no. We are a completely international company. Milan, for historical reasons, remains the biggest talent pool, but London, for example, is gaining momentum faster. Madrid. We will be hiring a lot of people in the States, I think, starting next year.

5. Building a tech giant out of Milan, Europe's talent pool & the outsider advantage

That raises an interesting question. Europe as a technology hub is not really what venture capitalists, even late-stage investors, are looking for. They look at the market there as maybe slower and maybe just not as good an opportunity. I think it's their decision: it's better to be in Silicon Valley or with American companies, or maybe in Asia.

So what's it like to be the most successful company in Europe, or one of them? Is there— I think you're right. No, I think it's—well, Spotify is obviously much better, but Klarna... You're probably in the top 10. Definitely in the top 10.

So, what kind of talent pool is there? How is it different, particularly in Italy? I noticed that when Chamath goes to Italy, there might be a little less work there: an additional button. It goes from 3 buttons to 4. Yes, the buttons are lowered, and the number of hours spent in front of a laptop is also decreasing. Chamath, we see it.

Do you keep these Italians in working order? How do you do it? What's the secret? Tell us about the talent in managing a company in Europe.

So, I think there are a lot of problems in Europe, but I think there is some pretty good talent. Half a billion people live in most of Europe, so that's a lot of people with a pretty good education. It's not Stanford, but it's solid.

Many of these people have some chance to prove that we are not necessarily less intelligent or capable. So, you will find a lot of good people. I think the fact that Italians don't work hard is mostly a false stereotype.

I find that—my wife doesn't work with me. She works in our company, and she works long hours. At our company, we work pretty hard. We typically find that when we acquire companies and work with existing teams, more often than not, the team we bring in is working significantly harder.

So, I don't know. We just try to hire people who are intrinsically motivated, very ambitious, and just want to be entrepreneurial. Then we give them a good reason to do their job the best they can, because they see that they can have a unique career anywhere other than Milan and that it can be a boost for the business.

New York is the banking capital, and Silicon Valley is the technology capital. Have you thought about moving your headquarters? Why? Luca's absolutely right. You come across people like this. The problem with people who go to these typical places and typical schools is that they think they're geniuses. When you actually look—even just at AI—who are the main participants? They're not from MIT or Stanford, per se. They're from McGill, CMU, and Waterloo. This could very well be an advantage.

Yes, well, that's what I'm getting at. This is a huge advantage.

I mean, when I interviewed Charles Koch, I talked about this. What impressed me was that he effectively built this business, perhaps the most unusual, entirely proprietary business on earth, from scratch in Wichita, Kansas. I say it's similar to the Wichita mentality because it essentially ignored everything traditional and could do things its own way.

I don't know anyone who thinks and does things the way you do and is based in Silicon Valley, and that might be because you live in Milan and aren't immersed in the cultural mindset.

Listen, I'll just say what I find incredibly interesting about your company and what you're building. Throughout the generations, we've seen these incredible examples of companies that used your blueprint, but in traditional industries: Amphenol, Roper, Danaher, Berkshire. We've never seen a successful implementation of this in technology, and I think you are the best large-scale example.

I mean, Expedia tried; Barry tried. I think it was a little difficult. So, it's really interesting to see that this thing can work, because the structural problem has always been: How do you guarantee these cash flows? And I think you're arguing that they're guaranteed—that these things can last 7, 8, 9 years.

Especially now, if you look at private equity, the private-equity guys are basically saying, “We don't know what the hell is going on,” right? However, you can still execute transactions and announce deals at the speed that many people in private equity do. So, how do you manage that risk? It is clear that this is not a risk. You think this is a tailwind for you?

Yes, I think private equity is very different because they keep these companies separate, mostly to sell them. They could never have that technology foundation, because once you plug it into a company, what do you do when you sell it to your private-equity competitor? Do you license it to them?

So, you remove that. They can't have a team of design engineers either, because if they put them on a business and then sell it, what do they do? They take the team, and without that right team, the business is worth next to nothing, or they sell the team along with it. So, the model is completely different, and I think those structural differences are the main reason why I would say we are successful.

So, it's never going to work with traditional private equity, which has other advantages. You can raise maybe a lot more capital because it's a little more hands-off, but you're never going to be able to achieve the kind of profitability that I think we have.

Okay, I'd like to thank you for the incredible business you're building. Congratulations.