Play Console acquisition reporting

This centre's own definitions are structured as a chain of visitor groups rather than a set of pure ratios, which makes them genuinely clearer than the other centre's version. And the headline number itself changed recently.

The definitions, in the centre's own words

People who visited your page, distinct visitors who opened your shop's own page but haven't actually installed it.

First-time installers, distinct visitors who installed your shop some stretch of time after opening your page.

People who kept it, installers who kept your shop on at least one of their own devices for however many days you're looking at.

Buyers: distinct visitors who bought at least one extra item, or had at least one membership charge, some stretch of time after opening your page.

Repeat buyers: distinct visitors who bought at least two extra items, or had at least two membership charges, after opening your page.

Conversion rate. On a chart, the percentage shown represents the move from one visitor group to the very next one in the chain.

Three things worth noticing

"People who visited your page" excludes anyone who already has it.

Read that definition again: visitors who haven't installed it. It isn't every single page visitor. This makes the visit-to-install ratio a genuinely cleaner measure of actual acquisition than the other centre's own version, and it means this number isn't directly comparable to the other centre's own page-view count.

"People who kept it" doesn't require anyone to have ever opened it.

It only means the shop stayed on at least one of their devices for that stretch of time. This is retention meaning "wasn't removed," not retention meaning "was genuinely used." A shop sitting untouched on someone's home screen for weeks still counts as kept.

That's a genuinely legitimate storefront-level number, and it isn't what a product team usually means by retention. Never put this centre's own "kept it" number next to an in-shop engagement curve without clearly saying which is which.

The whole chain moves group to group.

Conversion rate here is always the step between two neighbouring groups, so "conversion rate" on this side depends entirely on which two groups you actually mean. Always name them explicitly.

The main measurement itself changed recently

Store-listing reporting on this side shifted its focus from actual successful outcomes to raw visitor intent, specifically, clicks. The main number is now a specific kind of click: distinct visitors who clicked the install button right on your page. Older, acquisition-based reporting is still available separately in the statistics and export tools.

A click on install is not an install. It happens before the actual download, and that download can still fail, get cancelled, or come from someone who already has the shop. So click-based rates sit structurally higher than the acquisition-based ones they replaced, and:

• Any trend that crosses the point of that change shows a step that's really just an artefact of the measurement changing, not a real shift.
• Any older benchmark predates it entirely.
• "Conversion rate" is now genuinely ambiguous on this side unless you state exactly which number you mean.

Announced, but not yet actually documented

The centre has separately described further reporting changes at one of its own developer events that, as of this writing, don't yet appear in its own help documentation at all. Worth recording so the announcement is on file, and explicitly not treated as something you can actually use yet:

• an overall reach number for total visibility across the centre
• store-listing performance including value that wasn't previously reported at all
• traffic-source breakdowns extending downstream into engagement, retention and spending
• conversion rates specifically for extra purchases, included in the core performance numbers

Why this matters more than a typical roadmap footnote. A traffic-source breakdown reaching all the way into retention would answer a question this whole guide currently says can't be answered from the back-office alone, which source of visitors actually brings people who stick around. Worth checking for periodically. Don't build a report around any of it until it actually shows up in the back-office and in the real documentation.

Where traffic came from, and what people asked for

Traffic gets split into this centre's own search, its own browsing, ads and referrals, and an all-sources view. And the centre reports the actual words people searched before reaching your page, with anything asked for too rarely grouped into a catch-all bucket.

The search-versus-browsing split here is this centre's own equivalent of the other centre's search-versus-browse distinction, and the report of actual search words is free, genuine, unpaid query data that the other centre simply doesn't offer at all.

What tends to go wrong

Treating "people who kept it" as genuine engagement retention. It only means it wasn't removed.

Saying "conversion rate" without naming which two groups you mean. It's always a step between two specific ones.

Comparing across the point where the measurement changed. A genuinely different number.

Comparing "people who visited your page" to the other centre's own page-view count. This side's number excludes people who already have it; the other centre's doesn't.

Reading the catch-all "other" bucket of search words as pure noise. It's the long, specific tail, grouped together only because each individual word is small on its own.