Airbnb: Turning Off the Performance Tap in Public

In 2020 Airbnb cut most of its marketing spend because it had to, and found that the traffic largely stayed. It reorganised around brand marketing, said so repeatedly in public, and turned a cost decision into an argument the industry still quotes. Our verdict: PASS, with an asterisk about who may copy it.

The docking

The pandemic hit travel harder than almost any other category. Airbnb's bookings collapsed in early 2020, the company cut a significant share of its workforce, and its planned listing was in doubt. Marketing was among the first budgets suspended, including most paid search and paid social.

That is the context people forget when they quote this case. The initial decision was not a visionary repositioning. It was triage. The company paused performance marketing because it was burning cash on demand that had evaporated.

What happened next is the part Airbnb's leadership has described consistently since: when paid spend was switched off, direct and organic traffic held up far better than the attribution models implied. Brian Chesky has spoken about this repeatedly in interviews and on earnings calls, and shareholder communications have described a deliberate shift of investment towards brand marketing and public relations.

The implication was uncomfortable for the whole industry. A meaningful share of paid spend had been buying demand that already existed, and taking credit for it.

The manoeuvre

Airbnb rebuilt its marketing model around three things.

First, full funnel brand investment. Television, out of home and large brand campaigns returned, with the stated purpose of keeping Airbnb a verb rather than a search result. The 2021 hosting campaign, and later work promoting Airbnb Categories, were brand statements rather than conversion mechanics.

Second, public relations as a primary channel. Airbnb treats product announcements as press moments, with staged annual releases designed to generate editorial coverage rather than paid impressions, and credits PR with a substantial share of its reach.

Third, product as marketing. Categories, the host tools, the search experience and the widely discussed AirCover guarantee were built to be talked about. When the product is the news, the media budget does less work.

Alongside that, the company cut its dependence on paid search for branded terms, the most wasted line in large digital budgets: paying to appear against a query where the user already typed your name.

The result, as reported in its own filings, was marketing spend growing more slowly than revenue, with more traffic arriving direct and unpaid.

What worked

The experiment produced genuine evidence. Most organisations never learn which part of their paid spend is incremental, because switching it off is politically impossible. Airbnb was forced into the test and had the discipline to read the result honestly rather than restore the budget as soon as it could.

The strategic logic holds. For a brand with high consideration and verb status, performance marketing increasingly competes with itself. You are bidding to intercept people who were already coming. Moving that money upstream to create new demand is correct allocation, and it is the central argument of serious brand strategy work.

Treating product launches as earned media moments is efficient and durable. Press coverage carries third party credibility that paid media cannot buy, and a staged annual release calendar gives the newsroom a reason to pay attention repeatedly. It is a harder discipline than it looks, because it requires the product roadmap and the communications calendar to be genuinely aligned.

And telling the story publicly was itself a brand act. By making the argument openly, Airbnb positioned itself as a company with a point of view about marketing, which attracted talent and generated years of free commentary. There is a reason this case is in every conference deck.

What did not

The asterisk is substantial, and Airbnb has generally been careful not to oversell the case. Its imitators have not been.

Airbnb could do this because Airbnb was already a verb. It had a decade of category defining growth, enormous unaided awareness, an installed base of hosts and guests, a strong search position built on millions of listing pages, and genuine network effects. A brand in that position is not choosing between brand and performance. It is choosing whether to keep paying to harvest demand it already created. Most businesses have not created that demand, and for them performance marketing is not a tax on existing demand, it is the only demand they have.

Second, the claim is harder to verify than the retelling suggests. Traffic held up during a period when consumer behaviour, competitor spending, travel supply and search patterns were all moving violently. Isolating the effect of one company's spend decision in that environment is not straightforward, and the public case rests substantially on Airbnb's own framing of its own numbers.

Third, the brand led model has not solved the brand's actual reputational problems. Airbnb's hardest issues are regulatory and civic: housing supply pressure in constrained cities, local restrictions, fee resentment, and the gap between listing photographs and arrival reality. Brand advertising addresses the last of those weakly and the others not at all. A beautiful campaign does not change a city council's position.

Fourth, the organisational cost is real. This model requires strong in house creative capability, a disciplined PR function, and leadership willing to accept slower feedback loops. Many companies that announced a similar shift quietly restored their paid budgets once a quarter went badly.

The verdict

PASS, awarded for honesty about evidence rather than for the reallocation itself.

What earns the seal is that Airbnb ran an unplanned experiment, read the result against its own prior assumptions, acted on it, and published the reasoning. That sequence is rare. The common corporate behaviour is to rationalise the existing budget, protect the dashboard that justifies it, and never test the counterfactual.

The reallocation is correct for Airbnb and dangerous as general doctrine. The useful lesson is not that brand beats performance. It is that attribution systematically over credits the last paid touch, and that any brand with meaningful unaided awareness is probably paying twice for part of its demand. That is a measurement insight demanding a measurement response: holdout tests, geographic splits, incrementality studies. The right performance marketing programme knows which part of itself is incremental.

For a growing UAE business with low awareness, copying Airbnb's allocation would be close to negligent. Copying its willingness to test whether the spend works would be among the most valuable things it could do this year.

What a UAE brand should take from this

1. Run a holdout before you reallocate. Switch paid off in one emirate or one audience segment for a defined period and measure total demand, not channel attributed conversions.
2. Check what you pay for your own name. Branded search is the most common hidden subsidy in a digital budget. Test pausing it and watch total clicks, not paid clicks.
3. Earn your awareness before you economise on it. Airbnb cut performance because it had already built the demand. If nobody searches your name yet, performance marketing is your demand, and campaign management should reflect that.
4. Build a release calendar the press can rely on. Two or three staged announcements a year, with genuine news in them, beat a year of boosted posts on credibility and cost.
5. Do not expect brand work to solve a service problem. If the complaint is fees, regulation or an unreliable experience, fix the thing. Creative direction carries a good product further. It cannot carry a bad one.

Port Verdict also runs on our India desk at oldportindia.com. The full archive is at Port Verdict, and if someone has told you to cut performance spend because Airbnb did, please talk to us first.

Port Verdict is independent editorial commentary on publicly reported campaigns. Old Port Solutions has no affiliation with the brands analysed.