Brian Chesky on Airbnb and Product

Source: Lenny’s Podcast Speaker: Brian Chesky Source URL: https://www.lennyspodcast.com/brian-chesky/

Key ideas

• CEO as CPO: the founder/CEO should be the de facto chief product officer of a product company. Delegating product leadership takes founders away from their primary comparative advantage. The more Chesky delegated, the slower Airbnb moved — teams misread slowness as caused by over-involvement, creating a vicious cycle.
• Divisional company decay arc: advocacy → politics → dependencies → bureaucracy → complacency → stall. The functional model (design/engineering/marketing as shared functions, not divisions) is the antidote. Health heuristic: how close is engineering to marketing?
• Product-marketing integration: “you can’t build a product unless you know how to talk about it.” Airbnb merged PM and product marketing; the story of the product shapes the product. Semi-annual launches (May and November) package everything into a coherent narrative instead of a feature trickle.
• Leaders in the details: distinct from micromanagement. The board reviews the CEO without dictating; the CEO must review work without dictating. If you don’t know the details, you can’t evaluate performance or identify bottlenecks. One CEO review cadence = replaced political alignment with direct accountability.
• Add a zero: 10X thinking as a first-principles forcing function. The current process can’t scale 10X, so achieving it requires genuinely rethinking the problem from components. Also a leadership tool: seeing potential in people that they don’t see in themselves.

What happened with PM at Airbnb

Airbnb did not eliminate PM. It restructured the function:

• Merged inbound PM + outbound product marketing into one senior role called “product marketer.”
• Offloaded programme management to dedicated programme managers (previously embedded in PM roles).
• Made the group smaller and more senior — fewer junior PMs, higher responsibility per person.
• Removed control: product marketers manage by influence only. They do not own engineering or design resources.

Why designers at Figma Config cheered when they thought PM was eliminated: Silicon Valley often treats design as a service organisation — “catching things before they go out the door” — rather than as part of the development process. The frustration is real and structural, not personal.

The divisional company decay arc

A predictable failure mode in growing tech companies:

1. Separate teams drift onto different technical stacks → technical debt.
2. Everyone depends on shared services → bottlenecks (“backup like a deli”).
3. Teams give up waiting → build own versions of shared services.
4. Teams become self-sufficient divisions with separate roadmaps.
5. Divisions advocate for resources → politics (individual interests over company).
6. More specialisation → more boundaries → more interdependence → bureaucracy.
7. Hard to know who’s doing what → no accountability → complacency.
8. Company stalls. “10 marketing efforts, no customer heard anything.”

The result: marketing and engineering are in separate universes. Chesky’s heuristic — the health of a company is how close engineering and marketing are. At dysfunctional companies they’re in “different universes”; at healthy ones they’re “joined at the hip.”

The Airbnb reversal

Post-COVID restructuring (2020–2022):

Changes done in approximately this order:

1. Everything written down in a Google Sheet → dramatic reduction in active projects.
2. Functional reorganisation.
3. Fewer employees, fewer layers.
4. Expert leaders (no pure people managers).
5. CEO review cadence for all product and marketing.
6. Single shared roadmap + two launches/year.
7. Product marketing function created (PM + marketing merged).
8. In-house creative agency built.

CEO review cadence

Brian reviews all product and all marketing work on a rotating schedule: weekly, bi-weekly, monthly, or quarterly depending on the project’s phase and importance. A head of programme management scores every project red/yellow/green.

Why it works:

• If something isn’t happening, the review surfaces it instantly — no politics needed.
• Brian can feel “the equivalent of a semi-assembly of the entire new product every week” — like inspecting a car prototype to identify which individual component is blocked.
• No one can be a bottleneck without it becoming visible.
• Replaced all political alignment work: PMs don’t need to win people over; they need to get work reviewed.

The paradox of involvement: being in the details required significantly more work for 1–2 years. After the culture established, fewer conflicts, less turnover, fewer bad surprises (previously 9 of 10 surprises were bad; now 9 of 10 are good). The CEO now has more time despite being more involved in the details.

Product marketing: story first

“You can’t build a product unless you know how to talk about the product.”

Airbnb’s process: when working on a launch, one of the first things is to figure out the story. The story often dictates the product. A coherent story requires a coherent product.

This creates forced coherence: if you can’t tell a simple story, the product probably isn’t simple. If you can tell a clear story, the product is likely well-designed.

A/B testing critique: most AB testing is conducted without hypotheses. If A beats B and you don’t know why, you’re stuck with B forever. You can’t change it because you don’t understand it. Airbnb now requires a hypothesis for any experiment. A holdback (control vs. treatment) is acceptable; A/B testing without understanding is not.

Performance marketing vs. brand: performance marketing is a laser (precise, efficient, not compounding). Brand/product marketing is a chandelier (illuminates the whole room, builds long-term compounding advantage). The mistake: using a collection of lasers to light up an entire room.

Leaders in the details

“There’s this negative term called micromanagement. I think there’s a difference between micromanagement, which is telling people exactly what to do, and being in the details.”

Micromanagement = telling people what to do. Being in the details = knowing what’s happening to evaluate performance and hold accountability.

Analogy: the board reviews the CEO’s work without dictating it. That’s being in the details. Without knowing the details, there’s no way to evaluate whether people are doing a good job, which makes empowerment empty.

The paradox in delegation: the less Chesky was involved, the more spin there was, the less advocacy the team had, the fewer resources, the slower they moved. Teams assumed slowness was caused by over-involvement → they asked for more autonomy → more slowness → more delegation → cycle.

“Way too many founders apologize for how they want to run the company. They find some midpoint between how they want to run a company and how the people they lead want to run the company. That’s a good way to make everyone miserable. What everyone really wants is clarity.”

Add a zero

Airbnb’s internal phrase for 10X thinking. The purpose is not to mandate 10X outcomes — it’s a forcing function.

When you imagine something 10X bigger or 10X better:

• The current process is insufficient → you have to think differently.
• Thinking differently requires understanding the problem at a foundational level.
• Understanding at a foundational level = first-principles decomposition.

Leadership dimension: the John Wooden framing. A great coach sees potential in players they don’t see in themselves. Pushing a team isn’t saying “you’re not good enough” — it’s saying “I believe you have more in you.” The former creates demoralization (fixed mindset); the latter creates a growth mindset organisation where being pushed = being believed in.

Bias for action: “We don’t circle back next week. Let’s stay in this meeting until it’s done.” Decision speed = team pace.

Fake work and the reactive trap

Founders who only respond to inbound let their email and meeting requests set their agenda. The reactive mode means:

• Important relationships (friends, mentors) don’t reach out because “he’s busy.”
• The agenda is driven by whoever emails them, not by strategy.

Counter-practice: derive the year’s relationships and meetings from strategy. Ask: if life ends sooner than expected, who would you want to have spent time with?

“Fake work” = activities that feel productive but don’t move the ball. Meetings about meetings. Responding to things that don’t matter. Saying yes to things that are interesting but not important.

See also

• Brian Chesky
• Product Positioning
• Product Taste
• Organisational Kayfabe