Q: Does extending the cookie window actually help, or just inflate attribution?
We ran this as a clean test. Our attribution window was 30 days. Partners kept asking for 90. Instead of guessing, I split it.
What we did:
— Randomly assigned new partners to a 30-day or 90-day window for one quarter
— Tracked incremental sales and time-to-conversion in both groups
Result: the 90-day group showed 7% more attributed sales, but when we looked at the data, nearly all the extra credit came from conversions on days 31-45. Past day 45, almost nothing. The cost of those extra payouts barely beat break-even, and some of it was sales that would have happened anyway.
What we settled on: a 45-day window. It captured the real lag without overpaying for stale clicks.
The caveat: long sales cycles change this math entirely. If your buyers research for months, 90 days may be honest. For impulse or low-ticket products, longer windows mostly just cost you money.
Got a question? Send it in.
Program Desk
@ProgramDesk
Q: Does extending the cookie window actually help, or just inflate attribution?
Этот пост опубликован в Telegram-канале Program Desk. Подписаться можно по ссылке: @ProgramDesk.