Build the Greatest Cathedral

Luca Ferrari is the co-founder and CEO of Bending Spoons, the Milan-based company that buys and runs digital products including Evernote, Vimeo, Meetup and AOL. He grew up in a northeast Italian town of fewer than 1,000 people, the son of hairdressers, studied engineering physics, and worked at McKinsey while building his first startup, Evertale, at night. That company failed in 2013. Bending Spoons started 13 years ago in Copenhagen, moved to Milan, and now runs at roughly $3 billion in annual revenue. These signals are from Luca Ferrari's interview with David Senra on Senra.

1. Talent Over Experience. At Evertale, the best person on a team of about 12 delivered easily 10 times what the median person did, and was one of the least experienced people there. Ferrari says experience goes stale fast: the way you wrote software in 2010 bears little resemblance to 2026. Experience can also make people worse if it taught them low standards or office politics. "Experience, we can give it to you," he says, so Bending Spoons hires for a good brain and a massive eagerness to excel.
2. Hiring as Quant Science. Graduates have too short a track record to judge by outcomes, so Bending Spoons scores over 100 signals per applicant, the way a quantitative hedge fund stacks weak predictors. GPA is an obvious one. A subtler one: candidates are polite to interviewers, but how they treat the support staff who schedule their interviews predicts on-the-job collaboration far better.
3. Centralized Hiring and Firing. Ferrari says team managers have the wrong incentives in hiring. They want relief now, so they take the first adequate candidate and favor experienced people who need no coaching. A central talent team, staffed with brilliant people, sees a huge sample across 50-plus businesses and can move a product manager applicant into a growth role. Managers get no bonuses or KPI-linked pay; Ferrari says tying pay to concrete targets "almost invariably" leads to inferior outcomes.
4. Everyone Has One Job. Ferrari tells new hires that everyone at Bending Spoons, starting with him, has the same job: helping the company succeed. That's why an unfilled role doesn't worry him. Very little work in a digital business is strictly necessary. If 100 tasks are possible and only 10 are ROI-positive, many companies do 40, and some still miss a few of the 10.
5. Saturating Capacity. Promising people should have "way more on their plate than feels even remotely comfortable." The math: if you choose well, adding an 11th task can only raise the return on what you pick. It also trains people early to handle an enormous set of possible work, selecting a few high-return items and ignoring the rest. Ferrari would rather test that on day seven than discover someone can't do it when they run a 100-person organization.
6. Extreme Ownership. Bending Spoons borrowed the term from Jocko Willink and screens hardest for it. People with near-genius IQs have failed there because the job was their third or fourth priority. Recruiters look for proof of sustained caring, such as heavy open-source contribution or a startup someone ground at for 3 years. Ferrari points to Rafael Nadal, whom experts told him wasn't a top-50 talent in his generation, and says an extreme owner has less than a 1% chance of failing to have an amazing career there.
7. Relentless Simplification. Organizations left alone keep adding teams, process steps, features and rules, and each new piece adds connections to all the others. Going from 3 to 4 parts can make a system 40% or 50% more complex. At Bending Spoons, anyone proposing new complexity carries the burden of proof; the person objecting only has to raise a flag. Staff are also told to question existing things, and to question them harder the longer they've been around.
8. Delete the Whole Thing. About a decade ago Bending Spoons dropped job titles entirely. An algorithm labels anyone with product manager reports a "product management lead," whether they have 2 reports or 200, and people can write whatever they like on LinkedIn. Across multiple thousands of employees, Ferrari says nobody has ever complained. He applies the same logic to products: cut a feature 2% of users touch and serve the other 98% better.
9. The Operating System. Bending Spoons runs more than 50 proprietary tools for payments, A/B tests, lifetime value prediction, recruiting and AI model routing, and acquired companies are "installed" on them. Evernote cost about $200 million, had roughly $90 million in revenue, ran at about break-even and employed around 350 people. A 50-person task force rebuilt it; within about 18 months it had 50 to 60 staff, and today about 20 people run it while shipping features at least 3 times faster. Vimeo, roughly 4 times larger, took about the same 50 to 60 people.
10. AI Removes the Handoff. Ferrari says humans are "insanely inefficient at exchanging information," and Bending Spoons builds AI tools to cut out the handoffs. Diagram, an in-house design tool, follows each product's design guidelines, writes the code and pushes variants straight into A/B tests, sometimes in 1% of the old time. Alt Spooner is a Slack agent with each employee's exact permissions. Evernote's general manager used it to check support data for a bug, find the root cause, write a fix and ping the lead engineer for review in about 3 minutes.
11. Operator Edge Wins Auctions. Ferrari says 99% of the company's effort goes into being the best operator and remarkably little into deal-making. If a business is worth more under Bending Spoons than anyone else, it can be the highest bidder and still earn excellent returns. Ferrari believes the Evernote bid was way above the next one, and he can't recall being outbid in 5 years. Offers rarely rise more than 5% or 10%, and he encourages sellers to shop the bid, because a convinced seller signs faster.
12. Logic Over Numbers. Bending Spoons makes well over $4 million in revenue per core employee and pays some of the highest salaries in its markets. A 2-month test of higher posted salaries showed only modest uplift. Ferrari raised pay anyway, because a reputation as a top payer compounds over years at universities and on job boards, which a short experiment can't measure. "Logic and rationality, they never fail you"; numbers are useful, he says, only when handled with skepticism.

Watch the full video at https://www.youtube.com/watch?v=MCbHnlpwZf0.