Compounding Through Operational Excellence
What this covers
Paul interviews Luca about how Bending Spoons buys and runs software businesses instead of launching products from scratch. The economic case comes first. Only about one in 1,000 startups reaches half a billion in revenue, and survivorship bias hides how poor the expected return on most startup effort is. Luca then describes the cycle: operate better than almost anyone, pay sellers an appealing price, collect cash, and redeploy it, which he calls compounding. Deals are judged on IRR, NPV and opportunity cost, with little thought for whether the market will like them. Paul frames this as a weighing machine rather than a voting machine.
The middle of the conversation is about people and organization. Luca argues that caring spreads through a team, so talent density matters, and that leaders must show extreme ownership. He admits that recruiting for a hands-on culture excludes some competent candidates. The technology platform is standardized across acquisitions within months, the company runs like an open-source community, and teams report through Slack in whatever format works, to avoid bureaucracy. He also argues that proprietary technology cannot simply be bought. He expects AI to make ownership easier to sustain, because smaller teams carry more responsibility, and expects the company's automation to look like science fiction within five years.
The more contested material concerns markets. Luca calls investors' focus on organic growth shallow, claiming inorganic growth scales further than over 99 percent of organic strategies. He lists the advantages of being public and describes how rational discounting by short-horizon investors can feed mispricing. He also concedes that returns will probably compress as more capital is deployed. He ends on the founders' own near-zero odds of success.
Bending Spoons' founder argues that acquisition-driven compounding works only when the acquirer is a demonstrably superior operator, built on talent density, extreme ownership, and proprietary technology developed slowly over time, and that public markets' fixation on organic growth misprices this model.
- Capital allocation is the easy part once the company is the best operator; operating excellence is what makes acquisitions compound
- Culture, talent, and technology cannot be bought or hired quickly and require years of accumulated knowledge
- Decentralized decision-making with a standardized toolkit is what lets the model scale
- Inorganic growth has a far larger addressable market than organic growth, so investors' organic-growth focus is largely misplaced
Buying established, shrinking or product-market-fit businesses scales and compounds faster than organic growth or new launches.
- An inorganic strategy scales further than more than 99 percent of organic strategies, since even very successful single-product companies top out at one or two billion in revenue, while Bending Spoons is already at about $4 billion with a long runway.
“I think an inorganic strategy scales more than 99% of organic strategies.”
- Luca would sign a guarantee to deploy five to seven billion dollars a year for twenty years into shrinking companies at 65 percent levered IRRs, since negative organic growth is not a problem for the model, and about 80 percent of acquired companies were growing at purchase.
“if I could now sign a document guaranteeing that we can buy shrinking companies forever at 65% IRR, where do I sign?”
- Most investors focus on organic growth because they have been trained on companies like Google and because serial acquirers are rare, so organic growth is a shallow lens; many sophisticated investors understand that organic growth is almost irrelevant for this model.
“Part of it is just that not all investors are uber brilliant, and there's a lot of pattern following.”
- Valuation equals a multiple times a factor like EBITDA, and because growth compounds exponentially, faster growth in the earnings base can more than offset a lower multiple, and faster growth can even raise the multiple.
“if by allocating capital more efficiently, the other factor, say the EBITDA, gets to grow much faster... even if the multiple is a little bit lower, you're much better off”
- Launching products from scratch rarely pays for a company at Bending Spoons scale, because only about one in 1,000 startups reaches half a billion in revenue and such companies usually take about ten years to get there.
“by the way, there's probably a one in 1,000 probability that it does”
- To avoid running out of targets in five or six years, Bending Spoons is deliberately expanding into new markets such as hardware, enterprise SaaS through Airtable, and other industries, because its operating model ports well but not perfectly.
“expanding the addressable market early so that we never find ourselves at a point of saturation”
- Much of the startup ecosystem consists of talented people investing large amounts of time in failed projects for little pay, so the expected value of that investment is poor, and survivorship selection bias hides it.
“I think there is a lot of hidden loss-making investment in the world of startups”
- Bending Spoons buys companies that have already achieved product-market fit, and because it believes it can operate them better, it can pay a price that looks high to others while still delivering good returns.
“Instead, we buy product-market fit companies that have already succeeded. And because we believe we can operate them better, we can offer a high price, what's perceived to be a high price, and still deliver good returns for our shareholders.”
- Bending Spoons prefers acquiring established businesses with moats over launching products from scratch, because entering existing segments rarely erodes incumbents, and radical startup launches have dismal base rates of success.
“we like to own businesses with some moats, with some staying power, precisely because these businesses already exist in certain segments”
- Launching new products internally would be more sensible for a company like Apple, with an established customer base and distribution channel, but for Bending Spoons' diversified portfolio a new completely different product would not be appealing in expected return terms.
“by the way, there's probably a one in 1,000 probability that it does”
Talent-dense, caring teams built incrementally determine nearly everything a company achieves.
- Luca would not change the company's bootstrapped background, because although $100 million could have helped, there are worse things than having someone throw that money at you.
“So I wouldn't change our background necessarily, just because it's been good for the most part, but there are worse things in life than having someone throw $100 million at you.”
- Not receiving large early funding forced the founders to be resourceful and frugal, qualities that are paramount in the long run and are easier to make part of DNA when a company starts with little for a long time.
“There's value in having less. First of all, it forces you to be resourceful and frugal. I believe that in the long run, those qualities are paramount.”
- Caring spreads through a team, so a high density of people who care pulls people with latent potential toward caring, while a team where few care pushes the middle group toward indifference, so hiring and retaining for care is essential.
“caring is contagious, but also negatively contagious. So if almost everybody cares, the one person who cares slightly... they'll probably care.”
- If a company has a sensible strategy, which most companies lack, then talent and culture determine nearly everything, and with talented people who care about the company and strategy the company is about 80 percent there.
“if you have talented people who care tremendously about their job, the company, and they're pursuing the implementation of a strategy that's likely to succeed, you are already 80% there.”
- Culture cannot be installed by hiring hundreds or thousands at once; it must be built by aligning people on principles incrementally, starting with one person and multiplying, so that a shared playbook persists.
“there's no way to hire 500 or 1,000 amazing people overnight and have that be a unit of common purpose, a uniform playbook. It's just a recipe for disaster.”
- A strong employer brand cannot be built overnight or through advertising; it grows over years as graduates recruit peers, so that two becomes four and four becomes eight, eventually producing very large applicant pools.
“It's not advertising. It's many years of some of the best students graduating at a certain university coming to work here and then telling their former fellow students”
- Luca's personal AI agent, called Alternative Spooner, can request temporary database access, run analyses, and deliver results to him, so he can stay informed across businesses without relying on data scientists or his own database access.
“Access will be automatically granted for, say, five minutes, the amount of time that the agent needs to review the data.”
- Although Luca is a fan of data, almost any data can be manipulated, and metrics like revenue per employee can be inflated by simply not hiring enough people.
“We could just not hire people, or fewer than we should, just to make it higher.”
- Bending Spoons jokes that it buys established companies and returns them to startup mode, using small talent-dense teams with massive mandates rather than managed divisions.
“We sometimes joke that we buy established companies and we bring them back to startup mode. And we really do.”
- Bending Spoons cultivates an impact-driven culture in which people are motivated when their work matters for customers or monetization, rather than by passion for a particular product.
“we try to have a culture that's not passion-driven, but more impact-driven.”
- No single upstream number captures whether Bending Spoons is doing a good job, because its main performance enablers, team quality and culture, are qualitative, so proxies such as application volume and new-hire performance trends must be used.
“there's probably no single number that would capture whether we're doing a good job, because I think a lot of the enablers of the strong performance we've had over... 13 years have been somewhat qualitative in nature”
- Pursuing something truly massive requires urgency and long timelines, with a rocket-ship-like slow takeoff; people seeking immediate gratification through banking, consulting, or private-equity bonuses are unlikely to become exceptional outliers 20 or 30 years later.
“You are basically building an engine, a rocket ship that will take you longer to take off. But once it does, it'll move 100 times the speed”
- Paul argues that a company should expect to reinvent itself almost completely every decade, and Luca responds that the basic principles of buying product-market fit, operational excellence, and shared operating systems have stayed constant while AI is now the biggest change.
“The basic principles of what we do have stayed roughly the same from the beginning”
Capital allocation follows owner return math, not market preferences or venture pressure, to compound free cash flow.
- Large venture-style capital would not have been inherently harmful, but pressure to achieve early or different results could have derailed the company's slow-building progress.
“the pressures to get to certain results earlier, or different types of results, from say a VC potentially could have derailed our progress”
- Luca expects that within ten years there will be a great buyback opportunity, and at some point the stock may become cheap enough that buying back shares is more attractive than the next acquisition.
“There will be a point where the stock is stupidly cheap. Maybe a buyback is even more appealing than the next acquisition.”
- Bending Spoons should not optimize its portfolio to please markets, because markets change their preferences and buying a business only because others like it leaves the owner stuck with it, while bought businesses can be sold later.
“it would be dangerous to try to please the markets for a bunch of reasons. One is the markets change their minds in time.”
- A company should go public a bit early, before failing to be public becomes a real problem, because waiting too long could leave growth on the table.
“you want to go public when it's not a problem yet that you're not public. So we try to do that a bit early.”
- Astute investors may discount a stock because they expect other market participants to undervalue it, a rational strategy for those who will not hold forever, creating a vicious cycle of mispricing.
“there is this vicious cycle where they think, well, I think many people out there in the market don't get it.”
- As the company deploys more capital, it will need to buy a larger share of available companies, which creates downward pressure on returns through reduced selectivity, and the company expects returns to compress somewhat over time.
“I fully expect that there will be some saturation and some deterioration on returns, maybe because you need to deploy more and therefore you can be less discerning.”
- Henry Singleton built Teledyne by aggregating assets as conglomerates were popular and later spun off businesses when markets favored focused value propositions, and compounded efficiently through both regimes.
“there was a decade or two where what at the time were called conglomerates... were very popular.”
- Acquisition decisions are made on the math of IRR, NPV per Spooner, and opportunity cost, and the company barely considers whether the market will like a deal, leaning toward the market-favored option only when projected returns are essentially equal.
“it's all about the math of IRR, NPV per Spooner, these KPIs of compounding and ways of accounting for opportunity cost. We barely, if ever, talk about, will the market like it?”
- There is a wide margin between current returns and the level where the company would stop acquiring and start paying dividends, so the model does not depend on razor-thin margins.
“There is a pretty wide margin between where we are or have been up to now and, let's say, where returns would start being, eh, do we want to do this or do we start paying dividends.”
- Luca's strategy is to stick to compounding free cash flows as efficiently as possible, accepting that markets may like the company more or less at different times, because it expects this approach to go as far and as fast as possible in the long run.
“So we think it's much wiser to stick to compounding free cash flows as efficiently as we can”
Decentralized authority and a hard-working culture preserve team ownership against centralized approval and rigid rules.
- Decision-making at Bending Spoons is extremely decentralized: business and platform teams have full mandates, and nearly all decisions, about 99 percent, are made by teams without executive approval, with pricing as a rare exception.
“99% of decisions, the teams are trusted to take them.”
- Centralized decision-making is slow, can lack context, and most importantly deprives people of ownership, because if someone else must approve important decisions it is hard to feel responsible, which destroys productivity daily.
“centralized decision-making has multiple flaws... it deprives people of that precious sense of ownership”
- Whether Bending Spoons is centralized depends on how centralization is defined, since decision authority is decentralized while technology standards are centralized.
“that's completely true or completely false depending on how you define it. Centralization is a broad concept and the devil is in the details.”
- Leaders must consistently demonstrate extreme ownership, since failing to do so as a manager is the greatest failure, and Luca would rather a leader make a bad decision while demonstrating ownership than the reverse.
“I'd rather have someone make a bad decision but consistently demonstrate extreme ownership than vice versa.”
- Bending Spoons standardizes its underlying technology platform across all acquired businesses within months of acquisition, and operates internally like an open-source community where improvements benefit everyone, while principles remain non-prescriptive.
“So the underlying technological foundation is the same for everybody... We are like an open source community internally.”
- Strict standardized rules and reporting formats are a way for central management to impose its decisions at scale, and they create bureaucracy, so Bending Spoons avoids them and asks teams to report through Slack in whatever way works.
“sometimes rules are just an enabler for centralization, meaning if you strongly standardize how people operate and report”
- Recruiting that promotes hands-on, hard-working culture will exclude some highly competent candidates, such as later-career professionals who want status without hard work, but attracts a higher density of people who buy into the model.
“you give up on some of those by saying, here, everybody's hands-on, starting with the CEO”
- Bending Spoons has invested heavily to make predicting future behavior of applicants a science, because qualitative traits are hard to assess, especially for people with short track records.
“So we have made it a science, really.”
- After acquisition Evernote went from about 350 team members to about 60 within 12 to 18 months and is now run by 20 people, who launched the biggest product update in five or six years after cleaning up the code base and adding AI automations.
“Evernote, when we bought it, had 350 team members approximately.”
- Smaller teams managing the same product surface automatically give each person more responsibility and higher average talent density, both of which breed ownership, so AI makes the ownership culture easier to sustain rather than harder.
“if we have smaller teams managing the same surface of product or business... we automatically get higher average levels of responsibility and higher talent density”
Excellent operations make acquisitions and cash redeployment compound, so operating metrics need careful reading.
- Revenue per Spooner rose from about $1 million three years ago to over $4 million annualized, but this metric can be gamed by simply hiring fewer people, so data must be handled with care.
“We went from about $1 million three years ago to over $4 million annualized these days.”
- The downstream metric that best captures operating effectiveness is earnings per share or free cash flow per share, which should approximate each other over time, while no single upstream number captures the qualitative enablers like team quality and culture.
“The downstream metric, probably the one that's closest to capturing whether we are being effective or not, would be earnings per share or free cash flow per share.”
- Once a company has built an outstanding operating platform, the capital allocation side, particularly acquisitions, is the relatively easy part, so most effort goes into becoming the best operator.
“the capital allocation side of things, at least when it comes to acquisitions, is the relatively easy part once you have that platform”
- AI increases each person's output so fewer people are needed to absorb the same number of companies, which helps scalability and allows more acquisitions in parallel or larger ones.
“you need fewer people to, say, absorb the same number of companies or larger companies, which makes a lot of challenges easier.”
- If a company can operate businesses better than almost anyone, it can offer sellers an appealing price and still deliver high cash returns, then redeploy that cash into new acquisitions, and this repeated cycle is compounding.
“if you can operate a business better than almost anybody else, if a business is better off as part of your group... then you should be able to offer a price to those owners that's appealing and at the same time deliver high returns, cash returns”
AI lets revenue grow far faster than headcount through automated agents, tools, and documentation.
- Over the next five or six years, AI is the big difference maker because it can cause productivity to spike through incremental efforts in a way not seen before, and the company expects its automation to look like science fiction within five years.
“frankly AI is right now for this phase the big difference maker.”
- Internal AI agents let executives request data, run analyses, and receive compiled results autonomously, so leaders can catch up on many businesses without depending on data scientists, which makes a centralized-looking model scalable.
“I will ask it to run that analysis, and then it'll go and automatically request access to the relevant database.”
- AI allows the company to grow revenue tenfold while increasing headcount only about threefold, and the company aims to reach that ratio.
“I suspect we can 10x revenue by increasing headcount 3x at most, something like that.”
- Bending Spoons' internal tool Diagram is embedded in its toolkit with full product, database, design-guideline, and approval context, so designers work faster and more people can design without coordination loss.
“when you ask the tool to design something, it needs zero context.”
- AI can make writing memos or describing experiments roughly ten times faster, and Bending Spoons is trying to automate documentation fully, with much of it already automated.
“Now writing a memo or describing, say, an experiment becomes ten times faster than before, potentially.”
Public status provides cheaper capital and acquisition currency that private rivals lack.
- Being public gives a serial acquirer access to more and cheaper debt, easier equity raises, the ability to pay sellers in equity with deferred tax consequences, and easier buybacks, though public status brings headaches.
“public companies in general, especially those who do well as a public company, tend to have access to more debt, cheaper debt.”
- A public company cannot choose its shareholders, but it can clearly articulate its strategy and conduct itself consistently, which over time attracts a growing share of aligned holders.
“When you're a public company, you don't really get to choose your shareholders.”
- A public company can issue equity to sellers to pay for acquisitions in ways that avoid immediate tax consequences for those sellers, and sellers generally do not want illiquid private equity, so this is harder for private companies.
“You can issue equity to the seller to pay for some acquisitions in ways that don't trigger immediate tax consequences”
Massive ambitions demand urgency and patience because capabilities and outcomes take years to develop.
- The founders believed they had close to zero probability of building one of the most successful companies of all time, but downside was modest because they could still find jobs as engineers, so ambition was worth the attempt and helped attract better people.
“We knew we had close to 0% probability of achieving that, but we also thought, okay, we're probably not going to starve anyway in life.”
- Aiming for something truly massive requires urgency and hard work plus patience on timelines, since the engine takes longer to launch but then moves far faster, and early measurement at one to three years makes the path look like trailing.
“you are basically building an engine, a rocket ship that will take you longer to take off. But once it does, it'll move 100 times the speed”
- If you copy what everyone else does you will likely achieve a mediocre outcome, so outstanding results require means that also stand out, which led the founders to explore many solutions across hiring, culture, technology, and marketing.
“if you do what everybody else is doing, you will fail to achieve that objective... if you look for a way to be outstanding in terms of your outcomes, then very likely the means will have to stand out too”
- Work deserves an insane sense of urgency, because if something is worth doing it is worth doing well and quickly, and this urgency can be aimed at slow-payoff objectives like employer brand and predictive hiring tools rather than short-term revenue.
“I just don't think you should be doing anything in life, unless it's leisure, without an insane sense of urgency”
- A founder seeking an exit in three to four years by following a tried playbook or VC guidance will not build a Berkshire Hathaway-like company, though an unconventional long-term path only provides a chance, not a guarantee.
“I don't think you get to build a Berkshire Hathaway that way.”
- Proprietary technology cannot be bought faster with money because the knowledge of what to build and how is developed in micro-increments over long periods of problem-solving, failure, and acquiring and transforming companies.
“It's just a massive amount of knowledge that's developed in micro increments over a long period of time and solving many, many problems.”
Excelling at relative performance across many areas wins competitions and benefits the world.
- In business, relative performance matters: if a company does most things well, scoring about 80 out of 100 while most companies score 20, it will win most competitions, analogous to a basketball team shooting 45 percent against opponents at 38 percent.
“if your team shoots at 45% on average and most other teams are at 38%, you're very likely to win the vast majority of matches.”
- Two separate professional investors told Luca that Bending Spoons has been the most efficient and fastest compounder in history among companies they knew, relative to scale, though Luca says he has not studied this enough to be sure.
“you have always been the most efficient, fastest compounder in history, at least all the companies they knew.”
- Luca believes running businesses exceptionally well benefits the world even though it causes disruptive, painful short-term consequences such as parting ways with many team members of acquired companies.
“I do believe that the world is a better place if businesses are run exceptionally well, even though there are sometimes disruptive, painful short-term consequences.”
- Following John Wooden's definition of success as having done 100 percent of what you could, Luca argues that even a known finite runway is worth pursuing fully, just as building Google was worthwhile even though search eventually saturates.
“his definition of success was you have done 100% of what you could.”
The mission is to inspire European investment and distribute opportunity more broadly.
- Luca's motivations are competitiveness in business, pathological love of learning through talented people and broad mandates, and building something that resets the bar for running businesses excellently, with the aim of inspiring investment in Europe and distributing opportunity more broadly.
“I believe that the world is a much better place if knowledge, wealth, opportunity are more broadly distributed.”
- Building a world-beating company rooted in Europe and Italy could inspire more companies and investment there, and a large gap between major tech hubs and other regions is a societal problem because stability requires inclusion and fair distribution of assets.
“when that gap becomes a vast ocean, it's a big problem, including for the people who live there, because ultimately there's no stability in the world without a level of inclusion”