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

Bending Spoons’ strategy is to buy technology companies it believes it can improve, bring them onto shared foundational technology and a common operating team, and make them more cash-generative over roughly a decade. It then reinvests the additional earnings, using leverage to acquire larger businesses and repeat the process. Luca Ferrari said the company had completed more than 50 acquisitions, perhaps around 60, including Airtable, AOL, Eventbrite and WeTransfer. He sees the model as a way for scaled companies that have not reached the stratosphere to keep building value rather than simply wind down or sell for parts.

Ferrari’s compounding illustration starts with a business producing 100 of EBITDA: raise that to 200, borrow at 3x leverage to buy more businesses, improve those, and repeat. He said the process could potentially compound 100-fold over time. That does not mean maximizing debt or assuming every acquisition succeeds; Bending Spoons currently keeps leverage around 2.4–2.5x and historically closer to 2x. Ferrari’s broader point is that a lower valuation multiple can still produce a larger outcome if cash flow grows enough. The company has favored long-term free-cash-flow compounding over actions that might have improved its valuation in the short term.

Ferrari argues that radical invention and incremental innovation require different strengths. Radical breakthroughs can create higher, faster peaks, but are unreliable and may have a losing median outcome; incremental improvements are more measurable and can compound across products and acquisitions. He recommends choosing a primary approach that fits the company’s overall strategy rather than switching opportunistically. He considers the claim that businesses are either “bonds” or “options” simplistic but directionally useful: a leader responsible for one company may avoid bets that could destroy it, while a venture portfolio of roughly 50 companies can absorb many failures if four or five become 100x winners.

AI may make software cheaper to build and put pressure on margins, but Ferrari does not think software companies are therefore doomed. He sees durable advantages more often in brand, user habits, switching costs, network effects and first-mover status than in technical complexity. He said five strong new graduates could build a WeTransfer product about 80% as good in three months, while Eventbrite’s more important draw is its audience and event listings, including 90 million monthly ticket-seekers. In contrast, a routing company’s user-history data can capture muddy paths and landmark photos that satellite imagery misses. Ferrari expects AI to change product discovery and shift demand, but believes most digital products will be affected at the margin rather than disappear overnight.

Bending Spoons has built more than 50 internal tools for running digital businesses, while still using general-purpose products such as Gmail, Slack and Google Drive. Ferrari said its own tools reduce vendor costs, fit its needs and integrate with each other; moving an acquired company to a new stack all at once can be faster than replacing tools one by one. The company also uses a Slack-based agent with access to tools and assets to perform tasks such as data analysis, subject to credential controls. At Evernote, a general manager used an agent to investigate bug reports, inspect code, propose a fix and request an engineering lead’s review. Ferrari called the result “a hundred times better” than what had been possible before. He said AI handles most support tickets, while managers focus on designing response processes and user experience. AI-written code rose from 0% two years ago to 94% the last time he checked; he thinks a good company should reach at least 70%.

Ferrari describes Bending Spoons’ debt posture as designed to withstand severe downside: the company says it could repay all its debt before maturity even in severely adverse scenarios, refinances most debt annually to keep maturities more than five years away, and hedges 100% of debt against interest-rate fluctuations. He cites diversification, profitability and cash conversion as safeguards, but warns that founders can underestimate debt risk and that diversification may not help when a shock makes assets behave similarly. More broadly, he urges founders to clarify whether they want near-term valuation, a sale or long-term cash generation before choosing a strategy. He also cautions that he is not well placed to advise when to quit: his previous startup, EverTale, spent three years trying to use AI to write diaries automatically without gaining paying users, then ran out of money after a Danish government tax-refund issue. For Ferrari, ambition and instinct can generate ideas, but logic and rationality should determine which ones to pursue; talent density matters, and is especially difficult to improve when it starts low.