The B2B estimate attributes 22.1% of recipients’ gross booking value to gift-card receipt, with a 95% confidence interval of 3.5% to 40.6%. A second measure estimates $0.545 in additional booking value per dollar of gift-card value redeemed. Neither measure is Airbnb’s net revenue or profit. [S02]

Measuring what would have happened without the card

The commercial question is harder than counting redemptions. A recipient might use a card for a trip already planned, book a trip that would otherwise go to another provider, or increase the size of a booking. The same face value can therefore produce different effects on a merchant’s business. Redemption volume alone cannot separate those possibilities.

The authors combine 2024 observational records from existing North American customers with a 2022 experiment involving 26,980 existing customers. Their method addresses incomplete knowledge of who holds a gift card before making a booking. Its interpretation depends on assumptions about comparable customers, unobserved differences and the transfer of experimental effects to the observational population. [S02]

That approach matters for programme managers because card possession can remain invisible until redemption. A report based only on customers who spend a balance may miss recipients who never book. The study is an attempt to estimate the missing comparison, rather than a direct measurement of every recipient’s alternative decision.

For B2B cards, the model divides the estimated additional value into 7.1% of recipient gross booking value from greater spending among customers who would have booked anyway, and 15.0% from additional bookings. Their 95% confidence intervals are 5.7% to 8.4% and -3.5% to 33.5%, respectively. The second component is therefore individually uncertain even though the total B2B estimate excludes zero. [S02]

A new Stanford and Airbnb preprint estimates that B2B gift cards generated additional booking value, while the overall results for three other sales channels remain statistically uncertain. Justin Whitehouse and colleagues first posted the research on arXiv on October 6. [S01] The B2B estimate attributes 22.1% of recipients’ gross booking value to gift-card receipt, with a 95% confidence interval of 3.5% to 40.6%. A second measure estimates $0.545 in additional booking value per dollar of gift-card value redeemed. Neither measure is Airbnb’s net revenue or profit. [S02] Measuring what would have happened without the card The commercial question is harder than counting redemptions. A recipient might use a card for a trip already planned, book a trip that would otherwise go to another provider, or increase the size of a booking. The same face value can therefore produce different effects on a merchant’s business. Redemption volume alone cannot separate those possibilities. The authors combine 2024 observational records from existing North American customers with a 2022 experiment involving 26,980 existing customers. Their method addresses incomplete knowledge of who holds a gift card before making a booking. Its interpretation depends on assumptions about comparable customers, unobserved differences and the transfer of experimental effects to the observational population. [S02] That approach matters for programme managers because card possession can remain invisible until redemption. A report based only on customers who spend a balance may miss recipients who never book. The study is an attempt to estimate the missing comparison, rather than a direct measurement of every recipient’s alternative decision. For B2B cards, the model divides the estimated additional value into 7.1% of recipient gross booking value from greater spending among customers who would have booked anyway, and 15.0% from additional bookings. Their 95% confidence intervals are 5.7% to 8.4% and -3.5% to 33.5%, respectively. The second component is therefore individually uncertain even though the total B2B estimate excludes zero. [S02]

Your free reader account