App Store Analytics: Impressions, Product Page Views and Conversion Rate

In App Store Connect, impressions count views of your app on the Today, Games, Apps and Search tabs for more than one second, product page views count views of your product page, and conversion rate divides total downloads and pre-orders by unique-device impressions, as of November 2026. Read the three as one funnel from discovery to download, split by source.

This post is for iOS developers who run their own listing and see these numbers in App Store Connect. It gives each definition from Apple's own metric table, shows where to find the numbers, and works one small example through the funnel.

What does each metric count?

Apple defines the App Store metrics in one reference table. The five rows below use Apple's words in short form. Totals count views. Unique rows count devices. Conversion uses unique impressions only.

Two notes shape how you read the table. First, impressions already include product page views, so never add the two together. Second, App Store metrics appear once your app has at least five first-time downloads or pre-orders. A new app shows an empty state, not zeros.

First-time downloads mean first-time downloads on iOS, macOS, tvOS or visionOS. Usage totals beyond the store funnel reflect only users who opt in to share analytics with you. Apple states both limits in the same reference table.

Where do you find them in App Store Connect?

Open your app in App Store Connect, then open the Analytics tab. Under Acquisition, open Sources. The dashboard shows product page views by unique devices by default. Use the metric menu at the top to switch to impressions, conversion rate, downloads or other metrics.

Filter the view by territory and device to compare markets. Export larger sets through the Analytics reports API when you need bulk rows. Apple describes this flow on its Acquisition help page and points search readers to the same place to monitor impressions, conversion rate and downloads by source type.

If you run ads in search results, those views and downloads appear under App Store search as well. Apple includes them there by design. Keep this in mind before you credit a metadata change for a lift that paid placement drove.

How do you read them as one funnel?

Apple presents the funnel in this order: unique impressions to unique product page views to total downloads, with conversion rate linking the ends. Use unique rows for each step. Totals help you spot repeat views, but the funnel math needs devices.

The example below uses a clearly made-up app to show the math. Pixel Garden Tracker is not a real app. It had a quiet week with no featuring and no ads, as of November 2026.

Say the week shows 10,000 unique impressions, 2,000 unique product page views and 400 total downloads. The product page tap-through is 2,000 divided by 10,000, or 20%. The page-to-download rate is 400 divided by 2,000, or 20%. The conversion rate is 400 divided by 10,000, or 4%.

Now split the same week by source to find the weak step. Search brought 6,000 unique impressions and 300 downloads. Browse brought 4,000 unique impressions and 100 downloads. Search converts at 5%, browse at 2.5%. The listing is not the main problem. Discovery from browse is.

This split tells you what to change. Low tap-through from impression to product page points to the search result itself: name, subtitle, icon, ratings and first visuals. Low page-to-download with solid tap-through points to the product page: first sentence of the description, screenshot sizes and upload rules and whether reviews answer doubts. When ratings drag tap-through down, read how to reply to reviews before you rewrite anything.

Search traffic itself starts from text relevance for title, subtitle, keywords and primary category, plus behavior such as downloads and ratings and reviews. Apple lists these factors on its search page without weights. For the input you control, see how the 100-byte keyword field works.

Why does Apple show different numbers than estimate tools?

Third-party tools estimate downloads, revenue and keyword ranks from their own panels and models. App Store Connect reports what Apple counted on devices that share analytics. The two will differ. Neither side is wrong. They measure different things with different coverage.

Four Apple-side reasons explain most gaps. The store metrics need five first-time downloads or pre-orders before they appear. Usage-based rows cover opt-in users only. Unique rows count devices, while many tools report events or estimates. StoreKit loads count as product page views, so cross-promo inside other apps can lift views without a matching search or browse impression.

Apple also does not give keyword rankings, visibility scores or download estimates in App Store Connect. Those are tool metrics. Treat them as a vendor's model, not as Apple's count. When a tool and Apple disagree on downloads, trust Apple for what happened on the store and use the tool only for what it models.

What this doesn't cover

This post does not cover Google Play data. It does not explain sales, proceeds or subscription events beyond the download funnel. It does not rank estimate providers or judge their models. It does not tell you what a good conversion rate is, because Apple does not publish one. It does not document the Analytics reports API field by field.