Connect · 9 min · March 2026
What “product page conversion” actually counts
Teams paste a single percentage into a slide and call it App Analytics. The number is real. The denominator is usually not the one they think they are defending.
In App Store Connect, product page conversion is downloads divided by unique product page views, inside the date range and filters you left selected. That sentence is already enough to break a quarterly recap. Unique is not the same as total views. A person who opens your listing twice in a reporting window is not two opportunities.
Source type is the next quiet trap. Branded search visitors arrive already decided. They inflate the blended rate. Category search and browse traffic are the rooms where screenshot order and preview video still have work to do. If you only report the blend, a listing test on category keywords can look like noise.
Filters that change the story
Device class, territory, and source type should be named in every recap. A Thailand-only cut during a week that includes a public holiday is a different claim from a worldwide blend. Deveventgrid labs require the filter line on the recap page; we have failed write-ups that omit it even when the lift looked attractive.
Custom product pages add another denominator. Views on a custom page do not automatically roll into the mental model your PM has for “the listing.” Say which page you changed. Say whether the default page was left untouched. Otherwise you are comparing two rooms and calling it one conversion rate.
What this number cannot do
Product page conversion does not tell you why someone left. It does not isolate icon tests (those live earlier, at impression). It does not survive a UA spike without annotation. Treat it as a conversion surface metric, then go read impressions separately. That split is the whole point of App Store conversion insights as we teach them.
If you want the longer version with a sample export, the Conversion Lab spends week one on exactly this read.