Airbnb's New York bookings weren't stalled by its algorithm — they were stalled by bad photos, so the founders rented a camera and fixed the listings by hand, one apartment at a time.

问题

a marketplace's conversion problem is assumed to be a software or matching problem, when the real cause is upstream and unglamorous

背景

换别人会怎么做

The founders' first instinct, common in early-stage startups facing a conversion problem, was to look at the product and algorithm, better search ranking, better filtering, more prominent placement, treating a stalled marketplace as a matching or software problem to be solved with engineering changes.

他们看到了什么

The Airbnb founders saw, once they actually looked at the real listings customers were seeing, that the blocker wasn't matching or ranking at all, it was that most host photos were dim, poorly framed smartphone shots giving potential guests no confidence in what they'd actually be paying for, a problem no algorithmic improvement could touch since a listing that looked untrustworthy at a glance would fail to convert no matter how prominently it was surfaced. Rather than writing host guidelines asking for better photos and hoping hosts complied, they rented a professional camera themselves and manually photographed listings door-to-door, attacking the actual bottleneck directly instead of the systems built around it.

那一手

Rather than write host guidelines asking for better photos, the Airbnb founders rented a professional camera themselves and went door-to-door across New York City, personally photographing host listings to replace the poor-quality originals — a manual, unscalable intervention aimed directly at the actual bottleneck rather than the systems around it.

为什么管用

Personally rephotographing listings addressed the exact thing standing between a potential guest and booking confidence, trustworthy, high-quality images, rather than trying to route around bad photos with better algorithmic presentation of the same untrustworthy content. Because the founders diagnosed the real cause by physically examining actual listings rather than assuming the problem lived in the software layer, the fix went directly at what was actually suppressing conversion, which is why bookings on the rephotographed New York listings roughly doubled and revenue from the city doubled within about a month, a result no search-ranking tweak had achieved despite a healthy base of existing host listings. Doing the fix manually and unscalably first, rather than waiting to build a scalable photography solution, let the founders validate that photo quality really was the bottleneck before investing in the infrastructure to formalize it, which is exactly why the same approach then repeated successfully in Paris, London, Vancouver and Miami before becoming Airbnb's ongoing professional photography program.

值了多少

Bookings on the newly re-photographed New York listings roughly doubled, and the company's revenue from the city doubled within about a month of the effort. Airbnb repeated the same door-to-door professional photography approach in other key early markets including Paris, London, Vancouver and Miami with similarly strong results, before eventually formalizing it into Airbnb's ongoing professional photography program for hosts.

什么时候会失灵

The mechanism depends on the founders or team actually being willing and able to go look at the real product customers are seeing rather than only examining aggregate metrics or algorithm behavior, a team that never inspects the actual listings, code, or user-facing artifacts directly would miss an unglamorous root cause hiding in plain sight the way Airbnb's photos were. It also depends on the manual fix genuinely being something a small team can execute directly at meaningful scale for validation purposes, a bottleneck requiring specialized expertise or resources the founders themselves don't have access to couldn't be fixed by simply doing it themselves the way photography, a learnable skill, allowed. And an unscalable manual fix is only a starting point, not a permanent solution, since Airbnb still had to build a genuinely scalable photography program afterward to sustain the fix across every future market, meaning the manual intervention proves the diagnosis and buys initial results but doesn't substitute for eventually engineering a repeatable, scaled version of the same fix.

后来呢

The New York photo effort is one of the most frequently cited examples in startup literature of 'doing things that don't scale' — founders directly diagnosing and manually fixing an unglamorous root cause instead of reaching for a technical or marketing solution — and Airbnb's professional photography program, now offered to hosts at scale, traces its origin directly back to this manual intervention.

资料来源

1. [1]How Airbnb Used Pro Photography to Build an EmpireSnappr, 2022snappr.com
2. [2]Airbnb doing things that don't scaleAlexander Jarvis, 2020alexanderjarvis.com

收下它

同一路聪明

• 2012 · Buc-ee'sBuc-ee's turned the gas station bathroom — the part every competitor treats as pure cost — into the actual reason to stop there
• 2006 · Cedars-Sinai Medical CenterCedars-Sinai got doctors to wash their hands by photographing the bacteria growing on their own palms
• 1978 · William LeMessurier / CiticorpThe engineer who discovered his own finished skyscraper could collapse in an ordinary storm didn't hide it or panic publicly — he confessed to the client and had it secretly rebuilt around the people living inside it.