On March 25, 2026, App Store Connect Analytics got what Apple called its "biggest update since its launch" (Apple Developer News, March 25, 2026). It added more than 100 new metrics, cohorts, and two new peer benchmarks: download-to-paid conversion and proceeds per download. Four months later, Google rebuilt Play Console's store listing reports around clicks instead of installs, and added impressions across the whole store (Google Play Console, What's new, July 2026).
Both dashboards now hold more numbers than most teams can read. The names mislead, too. Apple's "conversion rate" isn't page conversion. Google's new CTR isn't its old conversion rate. And the benchmarks you've read online divide by something else again.
Below, you'll find what each metric counts and how to read it by source. Then come peer quartiles, cohorts, the Play changes and a monthly review that takes under an hour. This guide explains the numbers; when they point at the product page, the fixes live in our conversion rate playbook.
Key Takeaways
- Apple's March 25, 2026 update added 100+ metrics, cohorts and two peer benchmarks: Day 35 download-to-paid and proceeds per download (Apple, 2026).
- Apple's conversion rate divides downloads by impressions, not page views. Split it by source type before you judge the page.
- Read the conversion rate and download-to-paid quartiles together: the pair tells you who owns the fix.
- Google Play now reports CTR on clicks plus Reach impressions. Don't compare trends across July 10, 2026.
What changed in App Store Connect Analytics in 2026?
Apple's March 25, 2026 update added more than 100 metrics, mostly In-App Purchase, subscription and offer data. It also brought cohorts, two new peer benchmarks and up to seven filters at once (Apple Developer News, 2026). For ASO, the benchmarks and cohorts matter most.
Peer group benchmarks aren't new. Apple launched them on March 1, 2023, and they group apps "based on their App Store category, business model, and download volume" (Apple Developer News, 2023). That first set covered conversion, proceeds, crashes and retention. Apple previewed the payer metrics at WWDC25 and shipped them in March 2026, so benchmarks now reach past the install.
The redesigned dashboard has twelve sections, from Overview and Sources to Cohorts, Benchmarks and Metrics (Apple Help, Analytics dashboard, 2026). Four carry most of the ASO signal: Sources, Product Pages, Benchmarks and Cohorts.
| March 25, 2026 change | Detail | Why it matters for ASO |
|---|---|---|
| 100+ new metrics | In-App Purchase, subscription and offer data inside Analytics | Store funnel and revenue in one tool |
| Cohorts | By download date, download source, offer start date "and more" | See which source pays, not just which one installs |
| 2 new peer benchmarks | Day 35 download-to-paid and Day 35 proceeds per download (eight benchmarks in all) | Benchmark monetization, not only conversion |
| Up to 7 filters | "More than doubling what was previously available," per Apple's WWDC25 session | Source × territory × device reads without exports |
| 2 subscription reports | Exported through the Analytics Reports API | Join store data to your own warehouse |
Keyword data didn't change. As of September 28, 2026, search term isn't among the filters and dimensions Apple documents for Analytics (Apple Help, Filters and dimensions, 2026). Keyword-level data on iOS still comes from Apple Ads search-term reports, while Play Console, as you'll see below, does report search terms.
What does each funnel metric in App Store Connect Analytics count?
Apple's 2026 Help counts an impression when your app shows for more than a second on the Today, Games, Apps or Search tabs. Impressions include page views. A product page view is any view of your page, StoreKit loads included. Conversion rate is "total downloads and pre-orders divided by unique device impressions" (Apple Help, Metric definitions).
That last line matters most. Apple's conversion rate doesn't divide by page views. Four more definitions shape the dashboard.
- Total downloads are first-time downloads plus redownloads, counted "when a customer clicks the redownload button" (not auto-updates or restores). Conversion rate and both Day 35 benchmarks include redownloads, which flatters brand-heavy and re-engagement apps.
- Pre-orders count toward conversion rate once, and aren't counted again when the app downloads.
- Thresholds. Download metrics appear after five first-time downloads. Usage metrics need five active devices and cover only users who opt in to share data with developers. Analytics doesn't include watchOS.
- Source types. App Store Search "includes views and downloads from ads that appear in App Store search results." Browse covers the Today, Games and Apps tabs; App Referrer covers links and StoreKit pages in other apps. Then come Web Referrer, App Clip, Institutional Purchase and Unavailable (Apple Help, Acquisition, 2026).
Read that way, Apple's conversion rate is an impression-level metric that moves with ranking and placement mix as well as with page quality. Browse impressions from a featuring pull it down, while referral traffic pushes it up, because a web-link visitor lands straight on your page, and that page view is also the impression. That's our reading of Apple's definitions.
Why do conversion benchmarks disagree?
Each one divides by something different. AppTweak's 2025 US data puts the average App Store conversion rate, first-time downloads ÷ page views, at 8.56%. Its install rate, the same downloads ÷ impressions, is 3.8% (AppTweak, 2026). Same store, same year. The denominator alone opens a gap of about 2.3× (our arithmetic). The widely quoted "25%" is older still, AppTweak's first-half 2024 page-view figure via Adapty.
Same stores, different denominators (US, 2025)
The table below reconciles every conversion number founders quote, each from its owner's own definition, and rows only compare when numerator and denominator match.
| Metric | Numerator | Denominator | Window | Who's counted |
|---|---|---|---|---|
| Apple conversion rate | First-time downloads + redownloads + pre-orders | Unique device impressions (page views included) | Period you select | Your app, all or filtered sources |
| Apple peer conversion benchmark | Same as above | Same as above | Most recent week; 26-week trend | Peers by category, business model and download volume, with privacy noise |
| AppTweak conversion rate | First-time downloads | Product page views | Year (US 2025: 8.56% iOS, 16.15% Play) | AppTweak's benchmark data; method not disclosed |
| AppTweak install rate | First-time downloads | Impressions | Year (US 2025: 3.8% iOS) | Same |
| Google Play CTR (2026) | Unique users clicking Install, Open or Pre-register | Store listing visitors without the app installed | Daily or period you select | Your listing; peers at 25th, 50th, 75th percentile |
| Google Play conversion rate (retired) | Store listing acquisitions | Store listing visitors | Replaced mid-2026 | Superseded by CTR |
| Apple D35 download-to-paid | Downloads or redownloads followed by any In-App Purchase | First-time downloads + redownloads | 35 days from download | Your app and peer quartiles |
| RevenueCat D35 download-to-paid | Installs with at least one paid subscription | Installs | 35 days from install | 115,000+ apps on RevenueCat; medians |
Compare your App Store Connect conversion rate only with a benchmark built the same way: in practice, Apple's own peer benchmark, which uses the identical formula.
How do you read impressions, page views and downloads as one funnel?
Split Apple's conversion rate into two rates, by source type. Your search result card decides the page-view rate (unique page views ÷ unique impressions), and the full page decides page conversion (downloads ÷ unique page views). Start with search, since "almost 65% of downloads happen directly after a search" (Apple Ads, 2022 worldwide data).
Multiply the two and you get Apple's conversion rate back. In the illustration below, 30% × 30% = 9%.
One source, two rates: the App Store funnel (illustration)
There's a catch: downloads can bypass the page. Search results carry a Get button, so App Store Search downloads ÷ page views can pass 100%. A ratio above 100% isn't an error. It means the result card is doing the selling. Navigation apps show it best: AppTweak's first-half 2024 data, cited by Adapty, put them at a 115% "conversion rate" (Adapty, 2026). "Most users install navigation apps directly from search results without visiting the app page," Adapty explains.
Treat the icon, name, subtitle and first screenshots as a funnel step of their own, and see how your search result card renders before you change the page behind it.
Two traps: traffic mix and Apple Ads
Say App Store Search brings 100,000 unique impressions and 9,000 downloads, and Browse brings 40,000 and 400. Blended conversion rate: 9,400 ÷ 140,000 = 6.7%. A featuring doubles Browse impressions to 80,000 at the same 1% rate. Downloads rise 4.3%, to 9,800, yet the conversion rate falls to 9,800 ÷ 180,000 = 5.4%, about 19% lower. The page didn't get worse. The traffic mix did. (Illustrative numbers; we ignore devices counted in both sources.)
Net out Apple Ads next. App Store Search includes search results ad views and downloads, so a new campaign can pass for an organic win. Estimate organic as Search downloads minus Apple Ads installs, and compare trends rather than absolutes, because the two systems attribute differently. Our alignment guide shows how to net Apple Ads installs out of App Store Search.
Once the funnel is split by source, each leak points to one owner.
| What you see | What it points to | Where the fix lives |
|---|---|---|
| Impressions flat or falling | Visibility: rankings, featuring, ads | Ranking and keyword work, not the page |
| Low page-view rate on Search | The search result card: icon, name, subtitle, first screenshots | The conversion playbook, lever by lever |
| Low page conversion on Search | The full page: screenshots, ratings, preview video | Same playbook, tested with Product Page Optimization |
| Low conversion on Web or App Referrer | Message match between the ad or landing page and the product page | Custom product pages by traffic source |
| Healthy rates, falling downloads | Demand or seasonality | Market and keyword research, not the page |
How do App Store peer group benchmarks work, and how should you read them?
Peer group benchmarks place your app against apps with the same primary category, business model and download volume. Each shows the 25th, 50th and 75th percentiles. The default view covers the most recent week, and the trend view 26 weeks. Apple adds differential-privacy noise and says benchmarks provide "directional insight, not an exact ranking" (Apple Help, 2026).
Eight benchmark metrics exist, depending on the app. Six are conversion rate, Day 1, 7 and 28 retention, crash rate, and proceeds per paying user; the other two, added in March 2026, are Day 35 download-to-paid conversion and Day 35 proceeds per download. Free apps don't see proceeds per paying user. Apple also includes its own apps "where the customer journey is directly comparable."
Choose the peer group on purpose. You can switch to a subcategory or secondary category. You can also compare with All Categories under the same business model, or refine by download volume in View Trends. A utility listed under Productivity can land among the wrong peers, so pick the group your users would compare you with, then keep it fixed.
Half the benchmarks rest on a subset of users. Benchmarks "only use app usage data from users who've agreed to share their app analytics with app developers," Apple says. By our count, that covers four of the eight (Day 1, 7 and 28 retention and crash rate), while conversion rate, proceeds per paying user and both Day 35 metrics come from App Store and purchase data. Expect smaller samples behind the retention half.
Read quartile moves over four weeks or more, not decimals in a single week. Crossing a band is a signal. A 0.3-point wobble inside one isn't. And compare your conversion rate only with Apple's peer conversion rate, the yardstick behind check E6 in our ASO audit checklist, never with a vendor page-view benchmark.
What is download-to-paid conversion, and what's a good benchmark?
Apple defines Day 35 (D35) Download to Paid Conversion as "the percentage of first-time downloads or redownloads that were followed by an In-App Purchase within 35 days of download" (Apple Help, 2026). Day 35 Proceeds per Download averages proceeds over that window. RevenueCat's 2026 medians: 10.7% for hard paywalls and 2.1% for freemium apps (RevenueCat, 2026).
Split by user location, the spread is just as wide. The median is 2.8% for users in North America (top quartile above 6.0%) and 0.7% for users in India and Southeast Asia. Globally, it's 2.0%. The report covers more than 115,000 apps and $16 billion in revenue (RevenueCat, State of Subscription Apps 2026).
Day 35 download-to-paid: median and top quartile
Don't expect Apple's number to match RevenueCat's. Apple counts any In-App Purchase, consumables included, plus redownloads, while RevenueCat counts installs with "at least one paid subscription within 35 days," on its own SDK. Use RevenueCat for orientation and Apple's peer quartile for the verdict.
Apple's WWDC25 demo walks through a sample app. In it, "3% of your overall users are turning into paying users after 35 days, which is between the 25th and 50th percentile of your peer group" (Apple, WWDC25 session 252). That's below the median: a reason to dig in, but no emergency.
ASO moves traffic quality here: keyword intent, and message match between page and paywall, while paywall design, pricing and trials belong to our guide on trial-to-paid on the product page. One derived number helps (our arithmetic): D35 proceeds per download ÷ D35 download-to-paid ≈ 35-day proceeds per payer. At an illustrative $0.60 and 3%, that's $20, an early floor for lifetime value.
Read the two quartiles together
Conversion rate alone can't tell you whether to fix the page or the promise. Set your conversion rate quartile beside your D35 download-to-paid quartile, for the same peer group and period, and you get four states. It's our house framework, not Apple's.
Conversion rate × download-to-paid: who owns the fix
Low conversion with healthy download-to-paid means the page undersells a product people pay for: a conversion rate optimization (CRO) job. High conversion with weak download-to-paid is the state teams miss. The page or the keywords are pulling in people who were never going to pay, so the fix is the message and the keyword intent, not another screenshot test.
How do cohorts connect ASO traffic to revenue?
Cohorts group users by the day or month they downloaded or subscribed, then track them over days (D1, D2…) or months (M1, M2…). Apple's four cohort metrics are Download to Paid, Proceeds per Download, Download to Offer Conversion and Subscription Retention. Up to seven filters can apply at once, including source type and territory (Apple Help, Cohorts, 2026).
Four uses matter for ASO:
- Source quality. Compare D35 download-to-paid for App Store Search, Browse and Web Referrer. Apple suggests filtering by source type "to understand if a particular source brings in more paying users to your app than other sources" (Apple, App Analytics, 2026). Search still includes ads.
- Before and after a listing change. Compare download cohorts from the four weeks either side of a metadata or screenshot release. It's directional, not causal, and seasonality muddies it. For causal answers, run Product Page Optimization tests.
- Territory after localization. Apple's own example: after expanding to a new region, "monitor how long it takes users in that region to make a purchase compared to other more established regions" (Apple Developer News, 2026).
- Custom product pages. The Product Pages section reports each page's views, downloads and conversion rate, plus "downstream metrics like subscription performance and sales," once a page has five first-time downloads (Apple Help, Custom Product Pages, 2026).
Day 35 proceeds per download is also the input your bid math was missing. It's first-party revenue per install over 35 days, exactly the revenue figure in our break-even bid math and the Apple Search Ads Calculator. One caveat, from our analysis: it blends all sources and includes redownloads, so check it against your mobile measurement partner (MMP) data before you set paid-search bids.
Cohorts show which traffic pays. They don't prove what ASO caused. Incrementality tests and payback models belong to our guide on measuring ASO ROI and incrementality.
What changed in Google Play Console's metrics in 2026?
Google moved Play Console's store listing reporting from installs to intent in mid-2026. It now counts unique users who click Install, Open or Pre-register, and a click-through rate "replaces the legacy 'Conversion Rate' metric" (Play Console Help, 2026). Google also added Reach impressions across the whole store (Google Play, July 2026).
The dates need care. Google announced a reach metric and traffic-source breakdowns for engagement, retention and monetization at I/O on May 19, 2026 (Android Developers Blog, 2026). Google's Help makes unique clicks primary "from June 2026"; the Play Console announcement is dated July 2026. Reach spans homepages, cross-sell clusters, editorial placements, search, promotions and store listings.
Then there's the trend break. Google says it "improved our visitor methodology, which resulted in a decrease of visits, on average," starting July 10, 2026. It now counts indirect clicks too, crediting each piece of content a user touches before converting. Fewer visitors and more clicks can lift CTR with no listing change (our reading). Mark July 10, 2026 on every chart. Don't compare CTR across it.
Store listing clicks include the listing page, the inline install overlay and a mini-detail page Google may show for closely related queries (Play Console Help, 2026). In September 2026, Google added Day 2, 3, 14 and 28 retention next to Day 7 (Google Play, September 2026). The new metrics come with Traffic Source as a dimension. The table lines the two consoles up.
| Stage | App Store Connect | Google Play Console (2026) |
|---|---|---|
| Exposure | Impressions: shown 1+ second on the Today, Games, Apps and Search tabs | Reach: total, unique device and unique user impressions across six surface types |
| Page | Product page views, including StoreKit loads | Store listing visitors who don't already have the app installed |
| Action | Completed downloads (first-time + redownloads) and pre-orders | Unique users clicking Install, Open or Pre-register (intent) |
| Headline rate | Conversion rate: downloads + pre-orders ÷ unique device impressions | CTR: clicks ÷ store listing visitors |
| Benchmark | 8 metrics at 25th, 50th, 75th percentile; category, business model, volume | CTR vs peers' median, 25th, 75th percentile; by country, language, install state or source |
| Keyword data | No search-term dimension documented | Search term: "the term that the user searched for before navigating to your Store Listing" |
| Retention | Day 1, 7 and 28 benchmarks (opted-in users) | Day 7, plus Day 2, 3, 14 and 28 since September 2026, by traffic source |
This section only covers how Play measures the listing. For what to change on it, see Play listing tactics and store listing experiments.
The monthly ASO analytics review, step by step
Apple refreshes peer benchmarks weekly and keeps 26 weeks in the trend view (Apple Help, 2026), so a monthly read adds four data points. We run these six steps in the same order every month. They're the process behind checks A1, E6 and E7 of our audit checklist, and they take under an hour.
- Export. We pull 90 days of impressions, product page views, downloads and conversion rate per store, by source and top-five territory.
- Funnel by source. We compute page-view rate and page conversion for App Store Search, Browse and the referrers, with Apple Ads netted out of Search.
- Two quartiles. We read conversion rate and D35 download-to-paid against the same peer group as last month, and place the app in the grid.
- One cohort. We check D35 download-to-paid by source type, or by territory after a localization launch this quarter.
- Play. We read Reach, visitors and CTR by traffic source, plus the top search terms, with July 10, 2026 marked on every trend.
- Log and decide. We date every metadata and creative change per store, then pick one lever for next month. If it's the page, it goes to the conversion playbook. If it's the paywall, it goes to the subscription guide.
We pick only one lever because two changes in the same month make next month's cohorts unreadable. The discipline is the point. Same order, same peer group, one change at a time.
Frequently asked questions
How does App Store Connect calculate conversion rate?
Apple divides total downloads plus pre-orders by unique device impressions, per its App Store Connect Help. Because the denominator is impressions rather than product page views, Apple's rate usually runs below vendor page-conversion benchmarks, such as AppTweak's 8.56% US App Store average for 2025. It can also fall when featuring adds browse impressions.
What's the difference between impressions and product page views?
An impression is your app shown for more than one second on the Today, Games, Apps or Search tabs. A product page view is someone opening your full page, including through StoreKit inside another app. Impressions include page views, and downloads can happen straight from search results through the Get button.
What are App Store peer group benchmarks?
They're Apple's comparison of your app with apps in the same category, business model and download volume tier. They show the 25th, 50th and 75th percentiles for the most recent week. Since March 25, 2026, they include Day 35 download-to-paid conversion and Day 35 proceeds per download, eight metrics in all.
What is a good download-to-paid conversion rate?
Start with your peer quartile in App Store Connect. For outside orientation, RevenueCat's 2026 report puts the median Day 35 download-to-paid at 10.7% for hard-paywall apps and 2.1% for freemium apps. By user location, the median is 2.8% for users in North America and 0.7% for users in India and Southeast Asia. The global median is 2.0%.
Does App Store Connect show which keywords people searched?
No. As of September 28, 2026, search term isn't among the filters and dimensions Apple documents for App Store Connect Analytics. On iOS, keyword-level data comes from Apple Ads search-term reports instead. Google Play Console does have a Search term dimension: the term a user searched for before reaching your store listing.
Why did my Google Play conversion rate change in July 2026?
Google replaced it. Store listing reports now count unique clicks on Install, Open and Pre-register, a click-through rate replaces the legacy conversion rate, and indirect clicks count too. Google also changed its visitor methodology, which lowered visits on average from July 10, 2026. Don't compare the metric across that date.
The bottom line
App Store Connect now shows revenue next to the store funnel. Play Console now shows reach. Neither helps until you know what each number divides by.
- Apple's conversion rate is downloads ÷ impressions. Split it by source before you judge the page.
- Two rates: page-view rate for the result card, page conversion for the page.
- Benchmarks: Apple's peer quartiles, a fixed peer group, moves read over weeks.
- The grid: conversion rate and D35 download-to-paid together name the owner of the fix.
- Play: Reach impressions, CTR on clicks, and a break at July 10, 2026.
Want a second pair of eyes? Book a free 30-minute call and we'll read your funnel and benchmarks with you, or start with the fixed-price ASO audit. When the numbers point at the page, that's our conversion rate optimization work.
