Airbnb — The Brand-Over-Performance Pivot

Module: F9 — Marketing Finance Type: Performance Marketing Reckoning Case Cross-references: F9-04 (CAC and the performance marketing illusion), F9-06 (the business case for marketing investment), F9-05 (SOV, ESOV and the share laws)

The Situation

In the spring of 2019, Airbnb was the most-watched private company in the world. The IPO was perpetually six months away. The hotel industry — long dismissive of the "couch-surfing app" — had spent years quietly building competing platforms. Wall Street analysts had concluded that Airbnb's path to profitability ran through one place: the Google search results page.

The numbers told the story. By the close of 2019, Airbnb was spending roughly $5bn in total operating costs against $4.8bn in revenue. Of that operating cost base, marketing was the single largest controllable line item at approximately $1.6bn — about 34 per cent of revenue. Inside that marketing line, performance marketing — meaning paid search, paid social, retargeting, and affiliate channels designed to capture in-market demand — accounted for more than 80 per cent of the spend. By management's own later admission, in 2019 Airbnb was paying Google somewhere between $1bn and $1.2bn a year to bid on its own brand-related search terms and on category terms like "vacation rental Paris" and "cabin near Asheville."

The performance marketing dashboards looked superb. CAC was reportedly under $20 per booking guest. Return on ad spend was, by Airbnb's internal models, somewhere between 4x and 6x. Cohort payback was almost immediate. By every metric the performance marketing team had been hired to optimise, Airbnb was a textbook case of efficient digital acquisition.

Brian Chesky, the co-founder and CEO, was suspicious. Chesky is a designer by training, not an engineer or a finance person, and he had spent a decade watching his company evolve from a folding-air-mattress experiment in a San Francisco apartment into a global accommodation marketplace. His instinct, expressed repeatedly to senior staff in 2018 and 2019, was that the brand had been built by something other than Google AdWords. Word of mouth. The category-defining brand name. The "Belong Anywhere" 2014 rebrand. The press coverage of "Airbnb Open" host conferences. The cultural ubiquity that meant when people thought "rent a stranger's flat for a weekend," they thought one word.

But the dashboards said otherwise. Every booking that came through a paid Google ad was being credited to the performance marketing channel. Every retargeted Facebook ad that appeared the day before a guest finalised a reservation was being booked as that channel's "incremental" win. Inside the marketing operations team, voices warning that Airbnb might be paying Google to claim credit for demand it had already created were drowned out by the comforting authority of the attribution model.

This was the situation Airbnb walked into in February 2020. A profitable, fast-growing, IPO-ready company with a marketing budget that no one could fully explain — and whose CEO suspected, but could not prove, that the largest line in that budget was buying him very little.

Then COVID-19 happened.

The Decision

In March 2020, the global travel industry experienced a 90 per cent collapse in roughly six weeks. Airbnb's bookings cratered. Cancellations exceeded new reservations for the first time in the company's history. The IPO was shelved. Chesky raised $2bn in emergency debt financing from Silver Lake and Sixth Street at brutal terms, and on 5 May 2020 announced that Airbnb would lay off roughly 1,900 employees — about 25 per cent of the workforce.

Inside the marketing org, the cuts were even deeper. Performance marketing spend, which had been running at well over $80m a month at the start of 2020, was paused almost entirely. The total marketing reduction in 2020 versus 2019 was approximately $541m, the bulk of it from paid digital channels. For most of the second and third quarters of 2020, Airbnb effectively went dark on Google AdWords and Facebook. The company simply could not afford to keep buying clicks for an experience nobody could legally book.

This is the moment most discussions of the Airbnb story stop. The interesting decision is what came next.

By early 2021, vaccines were rolling out. Domestic travel began to recover. Bookings returned faster than almost anyone had predicted — driven by the so-called "rural rebound" as urban dwellers fled cities for cabins and remote houses, the precise inventory Airbnb specialised in. The obvious move, the one every CFO playbook would have recommended, was to turn the performance marketing taps back on. Recapture demand. Re-engage the paid funnel. Defend share against Vrbo and Booking.com.

Chesky did not do this.

In a series of internal meetings throughout 2021, and then publicly in investor calls and press interviews, Chesky made the call. The performance marketing budget would not be restored. Instead, Airbnb would invest in brand. The "Made Possible by Hosts" campaign launched globally in February 2021 — a deliberately quiet, emotionally-pitched film series featuring real guest photographs from hosts' homes, narrated by those hosts. Then came the May 2022 product redesign and the accompanying brand refresh, including the introduction of "Airbnb Categories" — a structural reframing of the entire booking experience around what kind of place you wanted to stay in, rather than where it was.

The old performance marketing budget did not disappear. It moved. The roughly $1bn that had been going to Google and Meta each year was reallocated — partly to brand campaigns, partly to product investment, partly straight to the bottom line.

Chesky was explicit in interviews about what he was doing. Speaking to CNBC's Andrew Ross Sorkin in February 2023, he gave the line that has become the canonical text for the marketing finance case: "Most of our traffic is direct or unpaid. We don't need to buy that. Performance marketing was largely claiming credit for demand we already had."

This was not a marketing director speaking. It was the CEO of a publicly listed company, in an investor-facing interview, telling Wall Street that for years his company had been paying nine-figure sums to a search engine to take credit for demand the brand was generating organically. The implication for the entire performance marketing industry was uncomfortable, and a good deal of the subsequent attribution-modelling debate of 2023 and 2024 traces back to this single quote.

The Data

The financial outcome was the part that ended the argument.

Metric	2019 (pre-cut)	2022 (post-pivot)
Total revenue	$4.8bn	$8.4bn
Marketing spend	~$1.6bn	~$1.5bn
Marketing as % of revenue	~34%	~17%
Net income	($674m) loss	$1.9bn profit
Direct/unpaid traffic share	~60%	~90%
Performance marketing share of total marketing	>80%	<50%

Several things matter about this table.

First, revenue did not just recover — it grew 75 per cent above pre-pandemic levels by the end of 2022. This happened despite the fact that Airbnb was spending less on marketing in absolute terms than it had in 2019, and dramatically less as a percentage of revenue.

Second, the share of traffic arriving "direct or unpaid" rose from roughly 60 per cent in 2019 to roughly 90 per cent by 2022. This is the most diagnostic number in the entire case. If performance marketing had been generating genuinely incremental demand, cutting it should have produced a corresponding collapse in traffic. It did not. The traffic stayed. Which means that traffic was not being generated by the paid channels in the first place — it was being intercepted by them, on its way to a destination it would have reached anyway.

Third, marketing as a percentage of revenue dropped from approximately 34 per cent in 2019 to approximately 17 per cent in 2022 — a halving. For a company growing this fast, that is an extraordinary structural improvement in operating leverage. Every dollar of revenue was costing Airbnb half as much to generate.

Fourth, net income flipped from a $674m loss in 2019 to a $1.9bn profit in 2022. Some of that swing reflects the post-pandemic travel boom, but the marketing reallocation is responsible for a substantial portion of it. Roughly $500m to $700m of the operating margin improvement maps directly to marketing spend that was no longer being deployed but was no longer needed either.

The Wall Street response was telling. When Airbnb finally went public in December 2020 — six months after the worst of the pandemic — the IPO priced at $68 per share. By late 2021, with the brand-led strategy publicly underway, the stock peaked at $216. Performance marketing efficiency, once the centerpiece of Airbnb's pitch to investors, was no longer the story. The story had become operating leverage, brand strength, and the durability of organic demand.

The Marketing Finance Lesson

This case validates F9-04 ("CAC, payback, and the performance marketing illusion") more cleanly than any other in the F9 module. The lesson is uncomfortable for an entire generation of digital marketing practice: under last-click and even multi-touch attribution, performance marketing channels routinely take credit for demand they did not create. The credit-claiming is not malicious. It is built into the measurement architecture. A user types "airbnb paris" into Google. Airbnb has bid on its own brand term. The user clicks the ad. The booking is attributed to paid search. By every dashboard at every weekly marketing review, that booking was "won" by performance marketing.

But the user was going to type "airbnb paris" regardless. The booking was created upstream, by all the brand-building work that made Airbnb the noun in that user's head. Performance marketing intercepted the demand — it did not generate it. And the cost of that interception, in Airbnb's case, was approximately $1bn a year.

This is what Les Binet and Peter Field have been saying for fifteen years and what most of the industry has been unwilling to hear: the long-term effects of brand-building generate the short-term efficiency that performance marketing dashboards then take credit for. The two are not in opposition; they are sequential. But when the measurement system flatters the second one, marketing finance directors stop investing in the first.

The case also speaks directly to F9-06 ("the business case for marketing investment"). Chesky's decision was not made on the basis of an attribution model. The attribution model, after all, said performance marketing was profitable. It was made on the basis of a forced experiment: COVID had cut performance marketing for him, and the world did not end. Bookings recovered without paid clicks. The natural experiment revealed the truth. The lesson here is methodological: the most reliable way to learn whether your marketing spend is incremental is to stop spending it and watch what happens. In normal conditions, no CFO will ever authorise that experiment. COVID authorised it for Airbnb.

And the case speaks to F9-05 ("SOV, ESOV and the share laws") in a less obvious way. By cutting performance marketing and reinvesting in brand, Airbnb was implicitly running an extra share of voice strategy: at a moment when competitors like Booking.com and Vrbo were continuing to spend heavily on paid acquisition, Airbnb was instead building category-defining brand work like "Made Possible by Hosts." The relative SOV in brand-building media — the kind that builds long-term mental availability — shifted dramatically in Airbnb's favour during 2021 and 2022. Which is exactly what the Binet and Field models predict will produce future market share gains.

The synthesis

The trap in reading the Airbnb case is to conclude that performance marketing is wasteful and brand marketing is virtuous. This is not the lesson, and it would not be the evidence-based lesson. Chesky did not abolish performance marketing — he reduced it from over 80 per cent of the marketing budget to under 50 per cent, and the function still exists at Airbnb today. The pivot was not Either/Or. It was a recalibration of the Both/And.

The The synthesis is that for years Airbnb had been operating on an Either/Or in disguise. The marketing finance team thought they were running a balanced portfolio because they had a brand budget and a performance budget. But within that balanced portfolio, the performance side was systematically claiming credit for the brand side's work, which made the brand side look comparatively inefficient, which led to the brand side's budget being squeezed in favour of further performance spending, which deepened the credit-claiming, and so on. The two halves of the marketing function had become structurally adversarial because the measurement system rewarded one of them at the other's expense.

What Chesky did, in effect, was force the system back into a genuine Both/And by interrupting the feedback loop. Cut performance hard enough that brand can be measured cleanly. Restore brand investment so that the long-term demand-creation function is properly funded. Then — and only then — bring performance back at the level where it is actually generating incremental demand, which turned out to be far lower than the dashboards had been claiming.

The evidence-based Marketing Finance Director, the figure F9-08 builds toward, would have caught this earlier. They would have insisted on incrementality testing. They would have run holdout experiments in geographic markets where performance marketing was paused. They would have read the attribution model with suspicion rather than reverence. The lesson of Airbnb is not that one side of the Both/And is right. It is that when one side of the Both/And is allowed to mark its own homework, the And quietly becomes an Or — and the consequences only become visible when something forces an honest measurement.

Sources

• Airbnb 10-K filings, fiscal years 2019, 2020, 2021, 2022. Filed with the United States Securities and Exchange Commission.
• Airbnb S-1 registration statement, November 2020.
• Brian Chesky interview with Andrew Ross Sorkin, CNBC, February 2023.
• "Airbnb Spends Less on Marketing, Reaps Higher Profits," Wall Street Journal, May 2022.
• "Airbnb's Marketing Cuts Are Working," Bloomberg, November 2022.
• Les Binet, "What the Airbnb story tells us about brand and performance," LinkedIn B2B Institute commentary, 2023.
• Peter Field and Les Binet, "The Long and the Short of It: Balancing Short and Long-Term Marketing Strategies," IPA, 2013, and the subsequent updates through 2019.
• "Airbnb Categories" product launch press materials, May 2022.
• Mark Ritson, "What Airbnb taught us about performance marketing," Marketing Week, March 2023.