Airbnb cut marketing spend by $446M and came out bigger
In 2020 Airbnb cut sales and marketing by more than a quarter. By 2021 revenue was 25% above 2019 on roughly the same reduced budget. The filings show what that actually bought.
One vote per reader. You can change it. No account needed.
What the filings actually show
The numbers below come from Airbnb's own annual reports, not from a press summary. Sales and marketing is a single audited line in the income statement, so it can be compared year to year without interpretation.
| Year | Revenue | Sales & marketing | Marketing as % of revenue |
|---|---|---|---|
| 2019 | $4,805.2M | $1,621.5M | 33.7% |
| 2020 | $3,378.2M | $1,175.3M | 34.8% |
| 2021 | $5,991.8M | $1,186.3M | 19.8% |
The cut itself was $446.2M, or 27.5%. That is the number in the accounts. It is worth saying plainly, because the figure most often quoted elsewhere is different — see the note on sources at the end of this piece.
The part that matters is 2021, not 2020
Cutting marketing in a year when your product is unavailable proves nothing. Anyone would do it. The interesting year is the one after.
In 2021 Airbnb spent $1,186.3M on sales and marketing — $11M more than the crisis year, and still $435M less than 2019. On that budget it earned $5,991.8M, a quarter more revenue than it had ever earned before.
That is the whole story in one ratio. Every dollar of 2021 revenue carried 14 cents less marketing cost than every dollar of 2019 revenue. Nothing about the product changed enough in two years to explain that. What changed is what the marketing budget was being spent on.
What a cut like this is really testing
When a company pauses paid acquisition and revenue does not fall by a matching amount, it has run an accidental experiment. The spend that can be removed without consequence was not creating demand. It was collecting demand that already existed — people who would have arrived through search, memory or a friend, and who were intercepted by an ad on the way.
This is uncomfortable rather than obvious, because that spend still shows a positive return in most attribution tools. The ad gets credited with the sale. The filings, unlike the dashboards, only see the total.
What this does not prove
It does not prove that performance marketing does not work, and it does not transfer to a business without Airbnb's brand recognition. A company nobody has heard of that stops advertising simply disappears. The finding is narrower and more useful than the headline version: at a certain level of brand awareness, a large share of paid acquisition spend is buying traffic that was already yours.
It also came at a cost that does not appear in the marketing line. Revenue fell 29.7% in 2020. This was not a clever budget decision taken from a position of strength; it was a company reacting to demand disappearing. The lesson is available to you without the crisis.
Three things to take from it
1. Run the test on purpose, in one market. Turn paid acquisition off in a single region for eight weeks and measure total revenue, not attributed revenue. The result is the only honest measure of incrementality you will get.
2. Track marketing as a share of revenue, not as a budget. A budget that grows with revenue looks disciplined and hides everything. The ratio is what tells you whether efficiency is improving.
3. Do not read this as permission to cut. Airbnb had a brand and a category position most companies do not. Test first, at a scale where being wrong is survivable.
Questions readers asked
Did Airbnb really cut $1 billion in marketing?
Not on the audited sales and marketing line. That line fell from $1,621.5M in 2019 to $1,175.3M in 2020, a cut of $446.2M. Larger figures quoted elsewhere refer to narrower internal categories of brand and performance media, which are not separately reported in the filings. We use the audited number.
Does this mean performance marketing does not work?
No. It means a share of it at Airbnb was capturing demand that already existed. That share is different in every business, and the only way to know yours is a holdout test.
Was the traffic recovery caused by the marketing change?
No. Travel demand recovered for reasons that had nothing to do with Airbnb. What the filings support is narrower: revenue exceeded 2019 while marketing stayed well below it.
How would a smaller company test the same thing?
Pick one region or one channel, turn it off for at least eight weeks, and compare total revenue in that region against a region you left running. Attribution reports cannot answer this question.
Where sources disagree: Trade press coverage of this story commonly reports a cut of roughly $1bn, or a fall to around $545M of marketing spend in 2020. Those figures do not match the audited sales and marketing line in the 10-K, which fell from $1,621.5M to $1,175.3M. The smaller figures appear to describe narrower internal media categories that Airbnb does not separately disclose. We have used the filed number throughout and are flagging the difference rather than choosing the more dramatic one.
This analysis is based on publicly available filings and reporting, all linked above. It represents our editorial interpretation of those sources. We have no affiliation with, and receive no payment from, any company named in this piece. If you believe a figure here is wrong, tell us — corrections are made publicly and the article is marked as updated.
Put your budget and channels into the simulator and see the range published cases landed in.