Video summary
The Playbook on Buying and Running Companies Forever
Main summary
Key takeaways
Business model & long-term strategy (“forever” acquirer)
- Bending Spoons as an operating acquirer: described as ~25% private-equity-style and ~75% tech company, where acquisitions are the engine of growth.
- Owns and operates permanently: unlike typical PE timelines (sell in ~3–5–7 years), they buy from the balance sheet to own and operate forever.
- Deep operational interventions (often radical):
- Rewrite software / re-architect cloud infrastructure
- Launch features, redesign UI
- Optimize monetization and marketing
- Sometimes rebuild big chunks or the entirety of the organization
Vision & operating principles
- “Company is the product” / building an institution (Berkshire Hathaway archetype).
- Culture as a differentiator: become the “ultimate testing and training ground” for top talent.
- Employer brand = most important product:
- Jobs are positioned to attract incredibly talented, hungry professionals
- Promise: high “talent density” and extreme selectivity
- Europe-first ambition: started in Europe to prove large-scale global tech institutions can be built outside the US/China default.
Core “playbook” for finding & buying businesses
Acquisition criteria (consistently stated)
- Digital technology focus: stay within/near circle of competence; expand boundaries cautiously.
- Scale: target businesses that are meaningful enough to enable hands-on ownership.
- Example planning logic: ~5 acquisitions/year max, with ability to go super deep.
- Predictability of future performance: sophisticated modeling is needed to know where the business is going.
- Clear ability to improve meaningfully: they don’t want deals that are “perfect already.”
Early acquisitions and compounding
- Started with ~€40,000 seed capital after a failed prior startup (Ever tale).
- Consulting attempts failed (no real contracts except ~€10,000 for a small app).
- First acquisition: ~€10,000 iOS app (keyboard personalization); sold for ~€20,000 quickly.
- Learning loop: small deals → revenues €10k → €20k → €40k → €80k, then compounding.
Execution advantages of the “home office” (platform model)
What the central platform unlocks
- Negotiating leverage
- e.g., cloud infrastructure and advertising partners
- cited as a few percentage points of EBITDA margin improvement (useful, not “transformative”)
- Cross-business R&D pooling
- Central team reallocates R&D based on fleeting opportunities
- Hiring/coaching is slow; with a platform, capacity can move quickly and then withdraw
- Cross-business marketing & resourcing fluidity
- Talent edge
- Higher growth, variety, and “testing environment” attracts better talent than a single-product average company
- Heavy investment in AI-based talent prediction
- described as extremely expensive to justify for small single-company hiring
- 2025 hiring example: ~800,000 unique job applications received; ~250 hires (≈ 1 in 3,200)
Metrics & KPIs mentioned
- Revenue growth & scale
- Recalled: ~“half a million a year” → “a billion” (over ~10 years)
- Example near-term scale: ~1.3B revenue “this year,” with ~75% YoY compounding (past four years)
- Selectivity
- 250 hires from ~800,000 unique applications (2025)
- Operational outcomes (Evernote case)
- ~250 significant product improvements in ~2.5 years
- Notes sync time improved to <10% of original, “in some cases 1%”
- Retention: “all time high”
- Pricing: increased; retention still improved
- Workforce examples
- General managers: average run businesses of ~$50–100M revenue
- Often ~27–28 years old (per speaker claim)
- Monetization / LTV direction
- Free vs paid tier segmentation (Meetup example)
- “Maximization of user LTV” referenced as objective (no numeric LTV/CAC provided in subtitles)
Case studies & actionable lessons
Ever tale → seed funding → strategy shift to acquisition
- Ever tale (2010): early AI attempt failed to scale; product didn’t work well with early ML.
- VC liquidation preferences meant the team was left with ~€40k.
- Remaining proceeds used as seed capital for acquisition-based growth.
- Consulting + sales outreach (“cold calling,” discounts) failed for months/years; a resilience crisis was described.
Actionable lesson: acquisition strategy partly responds to uncertainty in “zero-to-one,” with belief that functional execution can be improved more predictably than luck-driven outcomes.
Evernote acquisition (mid/large scale transition)
- Motivation: the small-scale acquisition model was saturating; needed structured companies with management teams/institutional scale.
- Deal dynamics:
- They won the bid (reportedly ~50% more than the next best offer)
- Believed it was “win-win”
- Transformation outcomes:
- Team of functional experts created a roadmap
- Rebuilt core codebase/cloud; performance improved dramatically
- ~60% higher cost to end users on average (framed by speaker), yet retention increased
- “Customer satisfaction” improved vs previous era
Actionable lessons:
- Winning bid with discipline: pay fair but competitive; don’t overpay beyond the model.
- Product + infrastructure execution can justify pricing power without harming engaged users.
Meetup pricing & monetization segmentation
- Introduced a free tier so organizers can participate; increased monetization for advanced use cases.
- Lesson: increasing price isn’t the same as better segmentation, personalization, comms, and packaging to raise LTV.
AOL (legacy brand, “sounds dead but isn’t”)
- AOL described as still having tens of millions of active users and strong loyalty.
- Speaker claim: AOL is the top-5 most used email inbox in the Western world.
- Lesson: “legacy” doesn’t mean weak—verify with real user and economics data.
Pricing & investment model (how they decide what to pay)
Offer discipline framework (stated explicitly)
- Determine return as a function of price (how IRR/NPV changes with price).
- Maintain discipline on walk-away thresholds (don’t pay above expected returns / opportunity cost).
- Negotiation approach:
- Not “offer max immediately”
- Aim for fair, early, competitive offers
- Balance “too low vs too high” to build reputation and negotiation leverage
Modeling method
- Debate assumptions (inputs) without looking at output during discussion.
- Run Monte Carlo simulation to evaluate distributions of:
- IRR
- NPV
- Negotiation “truth” is the simulated distribution, not optimistic bias.
Financing & capital structure approach
- Early years: relied on reinvested earnings / free cash flow; described as “completely true” for first ~5 years.
- Then: used debt (commercial banks), with rough reference of ~3.5x EBITDA leverage on “a good day,” generally lower leverage using trailing EBITDA.
- Equity raised mostly for secondary transactions (employee/team liquidity):
- secondaries organized every ~18 months / 1–2 years after equity accumulation
- claimed ~10% dilution (order-of-magnitude)
- Prefer permanent capital mindset (Berkshire-style) over traditional funds:
- not just for permanence, but to reduce “unnatural” liquidation timing pressure.
AI perspective (high-level, tied to execution)
- Belief: AI is an accelerator of quality and efficiency, but doesn’t replace execution.
- They already invest in AI internally and with custom integrations / proprietary tech.
- Model risk view:
- With highly diversified business units (most at ~20% revenue or less), localized declines aren’t existential.
- Acquisition thesis with AI:
- expect the gap between “cutting edge” and “laggards” to widen.
Leadership & organization design tactics
- No variable pay / low incentive complexity:
- everyone paid fixed salary
- no variable bonus, no stock grants
- employees may optionally invest some cash pay into equity at a discount
- Incentives instead through:
- hiring high-integrity, pride-driven people
- respect, culture, alignment to the aggregate institution (not BU-level KPIs)
- Reinforcing culture with rituals:
- “State of the Spoon”: twice yearly internal keynote-style sharing successes + failures + lessons
- Annual company retreat (7–9 days) to build trust and bonds
Systems thinking: consensus & decision-making
- “Consensus is overrated and dangerous.”
- Early struggle acknowledged: leader dealt with friction/criticism.
- When conviction is clear:
- listen with intellectual honesty
- accept disagreement and proceed, even if it upsets some people
Key timelines referenced
- 2010: Ever tale started; early AI product attempt (described as too early to scale)
- 2013: seed capital to start Bending Spoons after Ever tale liquidation outcome
- ~2.5 years: Evernote transformation cadence (per speaker)
- Past ~10 years: growth narrative from small revenue to ~billion scale
- 2025: hiring/selectivity and application volumes described
- 2019: attempted acquisition of Grindr (failed after raising constraints; “almost won”)
Sources / presenters
- Luca (interviewer; name not fully provided in subtitles)
- Patrick (interviewee/speaker; main source discussing Bending Spoons strategy and operations)