What should I fix first to raise my App Store conversion rate?

Whatever the leaking stage points at. If product page views divided by impressions is low, the fix is the icon, name and subtitle, because that is all a searcher sees. If installs divided by page views is low, the fix is the first two screenshots, the promotional text and the rating. Start with the cheapest change in the leaking stage: promotional text and subtitle take minutes and need no new build.

How long until a product page change shows in the conversion rate?

App Store Connect data runs about two days behind, and daily numbers are noisy. Compare the seven days before the change with the seven days after, starting the after-window once the change is live and the lag has passed. Judge after that full week, not the next morning.

Does the app description affect conversion rate?

Barely. On the product page only the first lines of the description are visible before a tap on more, and most visitors decide from the screenshots and the rating above it. The promotional text, which sits above the description and can be changed without a release, matters more than the description below it.

FunnelHound / Guides / Conversion rate optimization

App Store conversion rate optimization: nine fixes, ranked by effort

By Emin, indie iOS developer building FunnelHound · Published 3 Nov 2026

Conversion rate optimization on the App Store is two separate jobs. Impressions to product page views is decided by the icon, name and subtitle. Page views to installs is decided by the first two screenshots, the promotional text and the rating. Find which of the two leaks, then work the list below from the top: the cheapest fixes need no release and often move the number more than a redesign.

Two leaks, two toolkits. Fixing screenshots when the icon is the problem changes nothing, and vice versa.

Step zero: which stage leaks

Open App Store Connect → Analytics → Metrics and read three numbers for the last 28 days: impressions, product page views, total downloads. Divide them into two rates. Page views ÷ impressions is your tap-through rate; anything under roughly 5% in search-heavy categories means leak one. Downloads ÷ page views is the page conversion; the Benchmarks view shows where your peers sit, and below their 25th percentile means leak two. One wrinkle: search results with a Get button let people install without ever opening the page, so page conversion can read above 100% for apps that win on brand searches. That is a good problem.

Apple's own "conversion rate" metric is downloads ÷ unique impressions, which blends both leaks into one figure. The category numbers for that blended rate are in the benchmarks guide; this page is about what to change once you know yours.

The nine fixes, cheapest first

Effort against impact for a typical indie listing. Your grid shifts with the leak: an icon problem moves number 5 into the top-left box.

1. Promotional text (leak 2, ten minutes, no release). 170 characters above the description, editable any time. Put the single strongest reason to install here, in plain words, and change it when a feature ships or a season turns. Measure: installs ÷ page views.
2. Subtitle (leak 1, needs a version submission). Thirty characters under the name in search results. Say what the app does, not how it feels. "Scan receipts, export to Excel" beats "Your finances, simplified". Measure: page views ÷ impressions for search traffic.
3. First two screenshots (leak 2, a day of design, needs a submission). They render in the search result card and at the top of the page; most people never swipe further. Caption each with an outcome in large type, show the real UI, and make the first one stand alone. Our screenshot guide has the layout rules.
4. Ratings prompt timing (leak 2, an afternoon of code). The star average sits right under the name. Ask with SKStoreReviewController after a success moment, never on launch, and the average climbs over months rather than days. Slow, but every later fix inherits it.
5. Icon (leak 1, design sprint plus a submission). The only visual in search results besides screenshot one. One shape, one colour, no text, legible at 60 pixels. Test it through fix 7 rather than guessing. Measure: tap-through rate.
6. App preview video (leak 2, a week of production). Autoplays muted at the top of the page and pushes screenshot one off-screen on some layouts, so it can lower conversion for apps whose first screenshot already sells. Add one only if the app's value needs motion to be understood.
7. Product page optimization test (either leak, a sprint of assets and three to four weeks of traffic). Apple's built-in A/B test for icon, screenshots and previews, up to three treatments, with the result expressed as conversion lift. The PPO guide covers sample sizes and how to read a non-result.
8. Custom product pages (leak 2 for paid and social traffic). Up to 35 variants of the page, each with its own URL, so an ad about feature X lands on screenshots about feature X. Worth it once you run ads or have several large referral sources; pointless for pure organic search.
9. Localized screenshots (leak 2 in non-English storefronts, days of work per language). Captions in the visitor's language move page conversion in markets where you already get impressions. Check Analytics → Territory first; localize where the impressions are, not where you wish they were.

What does not move the rate

The keyword field. Invisible to users. It changes who sees you, not whether they install.
The long description. Folded behind a "more" link below the screenshots. Fine for keywords on some storefronts, almost irrelevant for conversion.
What's New text. Read by existing users deciding whether to update, not by new visitors.
Price changes. They change conversion a lot, but on a paid app that is a pricing decision with its own trade-offs, not an optimization. Free apps show no price at all.

Proving a fix worked

Change one thing, write down the date, wait for Apple's two-day lag, then compare seven days after with seven days before on the rate of the stage you targeted. A subtitle change that lifts tap-through from 4.1% to 5.0% while page conversion stays flat is a win on leak one and says nothing about leak two. Two changes in the same week cannot be told apart afterwards, so the slowest part of conversion rate optimization is the waiting, and the tool that helps most is a calendar. The change-measurement guide walks through the before-and-after arithmetic.

FunnelHound was built around that loop: each change becomes a milestone on the funnel chart, the pills show page views as a share of impressions and installs as a share of page views, and the before-and-after comparison for each milestone is computed for you. The leak with the worst pill is the one the app tells you to fix next.

Know which leak you just fixed

Add the change as a milestone, let the two-day lag pass, and FunnelHound shows the stage rates before and after it on your iPhone. No spreadsheet, no squinting at a daily chart.

Get FunnelHound

Share this on X

Data notes: metric definitions follow Apple's App Store Connect App Analytics documentation. Effort estimates assume a solo developer. Promotional text limit 170 characters, subtitle 30, custom product pages up to 35, product page optimization up to three treatments, per Apple's App Store Connect Help as of 2026. The 100 → 8 → 3 funnel in the first figure is illustrative.