The two-stage ranking model

App Store search ranking works in two stages that are easy to conflate but operate independently:

  1. Eligibility: does your app appear at all for a given query? This is determined entirely by your metadata — whether the search term appears (or can be inferred) from your app name, subtitle, keyword field, or in-app purchase names. If the term isn't indexed, you don't appear regardless of how good your app is.
  2. Ranking: among all the apps that are eligible for a query, in what order does Apple show them? This is where install velocity, ratings, engagement, and other signals come in.

Most developers focus on ranking when they're actually blocked by eligibility. If you're not appearing for a term at all, no amount of rating improvement will change that — you need to add the term to your metadata first. Check eligibility before diagnosing a ranking problem.

Signal 1: Keyword relevance (metadata)

The foundation of ranking is whether and where a search term appears in your metadata. Apple weights the fields in this order:

Exact match matters. An app named "Habit Tracker: Daily Streaks" will rank more strongly for "habit tracker" than an app with "habit" in the keyword field and "tracker" in the subtitle — even though both are technically indexed for the phrase.

Apple also appears to factor in how many other apps are competing for the same term in the same field position. A term in your app name that few competitors have in their name is more valuable than a term that dozens of apps all put in their name.

For the complete keyword strategy — how to find terms, what goes in which field, and how to avoid wasting your 100 characters — see the App Store keywords guide.

Signal 2: Install velocity

Install velocity — how many users are installing your app per day, relative to your competitors — is one of the strongest non-metadata ranking signals. Apps that are being installed at a high rate for a given query tend to rise in results for that query. This creates a compounding effect: better ranking → more installs → better ranking.

Velocity is measured relative to competitors, not in absolute terms. A new app getting 50 installs a day in a low-competition niche can outrank an established app with fewer recent installs. This is one of the few ranking signals that genuinely favors new apps if they launch well.

Install velocity is also why launch day matters. A concentrated burst of installs on launch — from your existing audience, press coverage, or a Reddit post — signals strong demand to Apple's algorithm and can establish early ranking positions that compound over time. See the app launch strategy guide for how to engineer that burst.

Signal 3: Ratings and reviews

Apple's algorithm factors in both your average rating and the volume of recent ratings. The key word is recent — a 4.8-star average from two years ago carries less weight than a 4.5-star average from the last 30 days. Apple appears to weight recent ratings more heavily than historical ratings, which means actively soliciting reviews matters on an ongoing basis, not just at launch.

The mechanics: Apple's SKStoreReviewController.requestReview() API surfaces the native in-app rating prompt. You can call it up to three times per year per app version. Timing matters — the highest conversion comes from calling it after the user has completed a meaningful action (finished a task, hit a milestone, saved something they care about) rather than on a timer or immediately on launch.

A low rating (below 4.0) doesn't just hurt conversion — it actively suppresses ranking. Apple's algorithm treats persistently low-rated apps as quality signals and can reduce their search visibility. Getting ratings above 4.0 isn't optional; it's a ranking floor.

For the full playbook on timing and placement, see the guide to getting users to rate your app.

Signal 4: Engagement and retention

Apple has stated publicly that user engagement is a ranking factor — specifically, whether users who install your app actually use it. An app with high install counts but near-zero retention is treated differently from an app with the same install count and strong Day 7 retention.

The engagement signals Apple likely uses include session count per active device, retention rate (Day 1, Day 7, Day 30), and crash rate (apps with high crash rates lose ranking, since Apple weighs quality signals). These can't be directly optimized through ASO — they're product signals. But they mean that ranking sustainably requires building an app people actually use, not just one that converts well on the product page.

Low retention also has an indirect effect: it increases uninstall rate, which some ASO researchers believe Apple factors in negatively. An app that gets installed and immediately uninstalled at high rates signals poor product-search match.

What the algorithm doesn't weigh

Several things developers commonly optimize for turn out not to affect search ranking directly:

How screenshots connect to ranking

Screenshots don't directly affect ranking — but they affect conversion rate, which affects install velocity, which does affect ranking. The chain is:

Better screenshots → higher product page conversion rate → more installs from the same impression count → stronger install velocity signal → improved ranking position → more impressions → more installs

This compounding loop is why the first screenshot is often the highest-ROI change you can make to a listing that has decent keyword coverage but flat rankings. A conversion rate improvement from 15% to 25% on 1,000 weekly impressions means 100 additional installs per week — a meaningful velocity signal for most apps in most categories.

The first screenshot is what users see in search results before tapping through — the caption, background, and frame visible in the search result crop. Getting that right (outcome-focused caption in the top 40%, non-white background, Bold font) is what the App Store screenshot template guide covers. ezscreenshots exports the variant at the correct dimensions for a Product Page Optimization test.

The ranking timeline

After a metadata change (keyword field update, app name change), new ranking positions typically take 1–2 weeks to stabilize. Apple's index doesn't update in real time — it recrawls on a schedule. Don't make a keyword change and check rankings the next day; give it at least 10–14 days before evaluating the result.

After a rating improvement or install velocity spike, ranking changes can happen faster — sometimes within days — because these signals update more dynamically than metadata indexing.

Ranking checklist: (1) confirm the term is indexed — search App Store for the exact phrase and check if your app appears at all; (2) if indexed but ranking low, check install velocity and rating relative to the apps above you; (3) if not indexed, add the term to name, subtitle, or keyword field in your next release.

The ranking ceiling for new apps

For highly competitive terms ("productivity app", "meditation", "fitness tracker"), established apps with years of accumulated ratings and install history have an advantage that's difficult to overcome quickly. A new app targeting these terms directly will likely rank in the bottom half of results even with perfect metadata.

The practical strategy: target specific long-tail terms ("pomodoro timer for students", "sleep tracker no subscription") where you can rank in the top 5 immediately, build install velocity and ratings there, then gradually expand to broader terms as your app's rating volume and velocity history grows. The full ASO strategy guide walks through this progression in detail.

Rankings compound from conversion. Conversion starts with the first screenshot.

A better first screenshot raises your install rate from the same search impressions — which strengthens your velocity signal — which improves your ranking. ezscreenshots makes the variant in minutes. Free, no account needed.

Try ezscreenshots →

Summary