How Airbnb grew by spending less on performance marketing
In 2020 Airbnb paused performance ads and leaned on brand, and by its own filings, direct traffic returned to pre-pandemic levels anyway.
The decision: pull back on performance
In March 2020, Airbnb paused its performance marketing spend and formally adopted a strategy to shift its marketing mix toward brand marketing and away from performance. This is documented in Airbnb’s November 2020 S-1, meaning the pullback began in 2020, earlier than the widely cited 2021 statements about it.[1]
For context on the scale: Airbnb’s advertising expenses had climbed from $403.9M in 2017 to $712.6M in 2019. In the first nine months of 2020, brand-and-performance marketing dropped roughly 65% year over year (from $842.2M to $294.8M) as the company cut performance spend hard.[1]
Why it worked
The bet was that the brand was strong enough that demand didn’t depend on buying clicks. The proof is in Airbnb’s own words: despite a material pull-back in performance marketing and the pandemic’s collapse of travel, its direct and unpaid traffic returned to 2019 levels by the third quarter of 2020.[1]
The structural tell is the traffic mix. Direct and unpaid channels made up 77% of traffic in 2019 and roughly 91% in the first nine months of 2020, so paid performance channels were only ever a minority of how people reached Airbnb, and shrank further. The company had earned its demand.[1]
77% → ~91%
Direct & unpaid traffic share, 2019 → first nine months of 2020.[1]
$403.9M → $712.6M
Advertising expense, 2017 → 2019, then cut sharply in 2020.[1]
−65% YoY
Brand + performance marketing, first nine months of 2020 ($842.2M → $294.8M).[1]
Recovered anyway
Direct/unpaid traffic back to 2019 levels by Q3 2020 with performance largely off.[1]
The system now
Airbnb’s FY2021 10-K describes its marketing as three components: brand marketing, communications (press, policy, and influencers), and performance marketing. It states plainly that the strength of the brand plus its communications strategy lets it be less reliant on performance marketing.[2]
In practice, editorial PR moments and cadenced product news do the demand-generation work that paid used to carry. Note: the FY2021 10-K does not quantify the direct-traffic share. The 77%/91% figures come from the 2020 S-1.[2] [1]
Where our system maps on.
Airbnb is the clearest case that a strong brand lets you buy less traffic. The nuance we’d stress with any client: Airbnb could pull back because it had already earned direct demand. You earn the right to spend less. You don’t start there.
Measure earned demand
Track how much of your traffic is already direct/unpaid. That share is the dial that tells you how dependent you are on paid.
Build the demand engine
Stand up the brand and owned-media machine (content, PR-worthy product news, SEO) that generates demand performance ads can’t manufacture.
Shift the mix as demand grows
Move budget toward brand as direct demand rises, while keeping performance on to capture high-intent buyers, exactly the balance Airbnb’s own model describes.
Defend direct-traffic share
Treat direct-traffic share as a north-star of brand health, so growth compounds off owned demand instead of rented clicks.
Sources
- [1]Airbnb, Inc. Form S-1 (Nov 2020): traffic mix, ad expense, performance-marketing pullback · U.S. SEC / EDGAR
- [2]Airbnb, Inc. Form 10-K, FY2021: brand / communications / performance marketing strategy · U.S. SEC / EDGAR
Independent analysis by MKA Growth Partners based on the public sources above. We are not affiliated with, commissioned by, or endorsed by Airbnb; trademarks belong to their owners. Figures are as reported at the time of writing.
Want this teardown done on your market?
We run the same analysis on your category, competitors, and funnel, then build the plan.