Everything moves at once
Seasonality, featuring and paid campaigns move installs in the same weeks.
ASO incrementality · App Store and Google Play
Control markets, a paid search check and rules set in advance.
Test marketsControl marketsExtra installs
Four weeks of data per market, before any change.Live in the test markets only. Paid budgets unchanged.Control markets show what test markets would have done. The gap is the extra installs.
The short answer
Change the listing in 2 or 3 test markets, leave matched control markets untouched, and compare organic installs in the same weeks against a baseline. This is the control-market method Admiral Media uses to prove ASO impact to finance teams.
Updated Every result links to a published case study or our own dated research.
Original research, November 2026
On 2 November 2026 we read the public App Store listings of 150 top apps in 14 storefronts. A listing that lags the app's languages has room to show a gap.
Pairs with that gap, by storefront
All pairs 24%
Source: Admiral Media portfolio store benchmark, public App Store data collected 2 Nov 2026. 25 portfolios, 37 developer accounts, 150 top apps, 14 storefronts. A selected sample, not a random one. Custom product pages and tests are not visible in public data and are not counted. Full method and findings
The problem
Seasonality, featuring and paid campaigns move installs in the same weeks.
Brand ads can take installs that would have come anyway.
Criteria set after the data give finance no reason to trust it.
Two-minute self-check
Tick what is true for your account today. Nothing is sent anywhere.
Your score updates as you tick.
Book a meetingThe method
Set up before the first change ships, so the result cannot be fitted afterwards.
Evidence
Three ways to measure an ASO change, weakest first. We use the strongest each change allows.
One market against its own past.
NeuroNation: +93% App Store search downloads in Korea after localized metadata, read before and after.
Read the case studyTest markets against matched, unchanged markets.
Our 60 to 90 day pilots, rules agreed before week 5.
Two versions on live traffic, split by product page optimization or the same Apple Ads ad groups.
ImmoScout24: −43% cost per acquisition with a custom product page against the default page, same Apple Ads ad groups, 4 weeks.
Read the case studyBefore the test
A written charter, and a check of what your brand ads really add.
Example charter. An illustration, not a client test.
Brand search check
Worked example
70% of the paid brand installs would have come anyway. The campaign added about 300 installs a week.
Illustration, not client data. Apple now shows more ads in App Store search, so a competitor may take the slot when you pause. What changed in App Store search
Published client results
Each figure links to a published Admiral case study.
LowerProperty appCustom product pagesApple Ads“Admiral Media were extremely helpful and efficient in making Apple's Custom Product Pages work for us.”
Alexandra ZinnowLead PPC, Digital Marketing, ImmoScout24−43%cost per acquisitionAgainst the default page, same ad groups, 4 weeksRead the ImmoScout24 case study 
HigherBrain training appKorean App Store+93%App Store search downloadsBefore and after localized metadata in KoreaRead the NeuroNation case study 
LowerMeditation appApple AdsNew markets−38%cost per install+54% click-through rateRead the Petit BamBou case study Need an ASO number finance will accept? Plan the first test with a senior strategist.
Book a meetingThe pilot
Part of an ASO program of at least 6 months.
Clients and recognition
“Our clients consider us a long term growth partner. We sail the same boat as them.”
Run the change in 2 or 3 test markets and leave matched control markets untouched. Compare organic installs in the same weeks against a baseline, hold Apple Ads budgets steady, and test brand search separately. Agree the decision rule before the change ships, then report the extra installs with your retention and revenue data, so finance sees value, not rankings.
Most market tests need 60 to 90 days: about four weeks of baseline, then eight or nine weeks after the change, because the stores take time to index new metadata and rankings settle slowly. Creative tests on live traffic can finish sooner, often in two to six weeks, when the app has enough store traffic.
For creative, yes. Product page optimization on the App Store and store listing experiments on Google Play split live traffic between versions, which is the cleanest read. They do not cover search rankings, so changes to titles, keywords and localization still need the comparison against control markets.
Pause brand campaigns in one market for two to four weeks while other markets keep running. Then read total installs from brand searches, paid and organic together. If organic installs rise by most of what paid delivered, the campaign was mostly buying installs you would have had anyway. Watch who takes the ad slot while you pause.
Markets with similar installs, seasonality and paid spend that share no listing with the test markets. On the App Store one localization can serve several storefronts, so countries that share a language rarely control for each other. We pick controls from your own history and check that they moved with the test markets before the change.
Then the comparison runs against time and signals the change cannot touch, such as installs on the other store or branded search volume, over a long baseline. That is weaker evidence, so the report says so, and split tests on creative carry more of the proof.
Read access to App Store Connect and Play Console for impressions, product page views, conversion and installs by source, your Apple Ads reports, and retention and revenue by market from your attribution or BI tool. Access is role-based and approved by your Legal and InfoSec teams first.
No, it feeds them. A market test gives a measured lift for one change, which can calibrate a marketing mix model. Organic store installs carry no channel in attribution, so tests are how ASO earns its line in the model, and your attribution data supplies the retention and revenue we join to the extra installs.
Admiral Media does not publish prices. Measurement is part of an ASO program, scoped by the apps, markets and stores in it and how often you test. Programs run at least 6 months, and a one-time ASO audit is the other way to start.
Book a meeting
Tell us the app, the markets and the question finance is asking. A senior strategist replies within one business day.
What happens next
Start an enquiry
