How much conversion volume appears the moment you start counting view-throughs?

How much conversion volume appears the moment you start counting view-throughs?

A display campaign's reported conversions tripled after the team enabled view-through attribution — crediting a conversion to an ad that was merely served, never clicked. The question: did the campaign suddenly work, or did the counting rule just become generous?

The mechanism. View-through attribution credits an impression that was shown within a lookback window before a conversion, even with no interaction. Because display impressions are cheap and vast, the lookback can sweep in enormous numbers of conversions that any large impression volume would statistically "precede" — including bot impressions and ads served below the fold, never seen.

What controlled tests reveal. When brands have validated view-through claims against holdouts, the incremental fraction is often tiny. Documented cases show view-through attribution inflating reported conversions by 2-4x while true incremental lift, measured by a public-service-ad or ghost-ad control, was a small single-digit percentage. The wider the lookback window, the more credit appears — and the more of it is coincidence.

The nuance. View-through is the most direct case of mistaking temporal correlation for causation. "Ad shown, then conversion" is a sequence, not a cause. With enough impressions, you can "precede" almost any conversion.

Bottom line for practitioners: treat view-through conversions as an upper bound on a channel's contribution, never a measurement of it. If enabling view-through tripled your numbers, you changed the rule, not the world. Validate with a ghost-ad holdout before crediting impressions nobody clicked.