Why attribution windows matter
They’re the quiet reason numbers disagree: the same day’s sales, credited under different windows, produce different ROAS from the same reality, and platform defaults are chosen by parties with opinions. View-through credit especially inflates, counting purchases from people who merely scrolled past, which is how retargeting learns to claim everything. Reading any platform metric starts with asking which window wrote it.
Working with windows deliberately
- Know each platform’s default and what your account overrides
- Shorten to match reality: impulse categories close fast; long windows import coincidence
- Treat view-through skeptically: a seen ad is not a caused sale
- Compare like with like: cross-channel numbers aligned to one window before any verdict
Questions
Which attribution window is correct?
The one matching your purchase cycle, applied consistently: quick-decision products justify short click windows, considered purchases longer ones, and the choice matters less than holding it steady while comparing. Incrementality tests calibrate what any window merely asserts.
Why did results drop when the window shortened?
Nothing changed but the bookkeeping: conversions that used to be credited now fall outside the frame, and reported ROAS steps down while real sales hold. Window changes require restating history, or every trend chart lies at the seam.