What is a view-through conversion?
A view-through conversion is a conversion that is credited to an ad the person saw but never clicked. Google Ads Help defines it as a conversion that occurs after an ad impression where the user does not interact with the ad and then later converts, counted only if the conversion falls inside the view-through conversion window (the period after an impression during which such a conversion is recorded).
Every part of that definition matters. There was no click, so there is no direct evidence the person noticed the ad at all. The only thing recorded is that the ad was served in their browser or app, and that a conversion happened afterwards. Whether the first event caused the second is exactly the question the metric cannot answer on its own.
Why view-through credit is worth having
Dismissing view-through conversions entirely is as wrong as trusting them completely. Display, video, connected TV, and social feed placements do influence people who never click. Anyone who has watched a brand campaign lift branded search volume has seen a real view-through effect. If your reporting counted clicks only, upper-funnel activity would look like pure cost, and the budget would drain toward the bottom of the funnel until demand ran out.
The metric also matters more as click data gets thinner. With shorter cookie lifetimes, in-app browsers, and platform-level tracking limits, a growing share of genuine influence never produces a clean, attributable click. Impression signals are one of the few remaining ways to see it at all, which is part of the wider story in cookieless attribution.
Why view-through credit gets overstated
Three mechanics inflate the number in almost every account:
- The impression was cheap and everywhere. Programmatic and social inventory serves enormous volumes of impressions at low cost. When a large share of your likely buyers sees an ad at some point in a month, some of them will convert regardless, and a view-through window will claim those conversions.
- The audience was already primed. Retargeting pools consist of people who have already visited your site. Serving them an impression and then counting their purchase as view-through credit is close to circular.
- Viewability is not attention. An impression counted by an ad server means the ad was rendered, not that a person looked at it. Below-the-fold and briefly-visible placements still generate impressions that a view-through window will treat as influence.
The result is that view-through conversions are the single largest contributor to platforms collectively claiming more conversions than the business actually recorded.
How to handle them in practice
- Report click-through and view-through separately, always. A blended figure hides the assumption you are relying on. Two columns costs nothing and prevents most of the arguments.
- Keep the view window short. Attribution windows for impressions are typically much shorter than click windows for exactly this reason, and lengthening them adds coincidence faster than it adds insight.
- Deduplicate across platforms. Left alone, several platforms will each claim the same order on a view-through basis. Reconciling to one source of truth for revenue is the only way to get a total that sums correctly, which is what independent multi-touch attribution exists to do.
- Validate with a lift test, not an argument. The honest answer to “are these view-through conversions real?” comes from an incrementality test that holds a comparable group back from the impressions. That is the only method that separates influence from coincidence.
A reasonable position to take
Treat view-through conversions as a directional signal about upper-funnel contribution, not as revenue you can bank. Look at whether they move in step with things that are hard to fake: branded search volume, direct traffic, new-customer rate, and overall conversion rate during a burst of impression-heavy spend. When they do, the view-through effect is probably real. When paid impressions climb and every independent indicator stays flat, the metric is measuring coincidence, and the reported versus incremental ROAS gap is where that shows up.