How the App Store Algorithm Works in 2026: A Marketer's Guide
Apple's App Store ranking is not a black box. Eight signals decide which app shows up first. Here are the four you can actually influence, ranked by impact.
The App Store ranking algorithm scores every candidate app against the user's query using eight known signals. Three of them you can't influence directly. Four of them you can. One is set by your competition. Understanding which is which is the difference between an ASO strategy that compounds and one that wastes budget.
Here's the 2026 picture for both Apple's App Store and Google Play.
The eight signals that decide rank
Across both stores, the ranking inputs are:
| Signal | Apple | Influence-able | |
|---|---|---|---|
| Install velocity from query | Heavy | Heavy | Yes |
| Conversion rate | Medium | Medium | Yes |
| Star rating | Medium | Medium | Yes |
| Review volume + recency | Medium | Medium | Yes |
| Metadata relevance | Heavy | Heavy | Yes (one-time) |
| Engagement / retention | Medium | Medium | Indirect |
| Search Ads spend (paid lift) | Small | N/A | Yes (paid) |
| Competition in the query | Variable | Variable | No |
The first four are the ASO levers. The fifth (metadata) is a one-time setup. The sixth (engagement) you influence by building a better product. The seventh exists only on Apple. The eighth is just gravity.
How install velocity gets read
Install velocity from a query is the single biggest influence-able signal. The store doesn't care about your absolute install count; it cares about three things:
- Your share of installs from this query in the last 7 days, compared to apps in your rank band
- The trend in that share, week-over-week
- The pacing pattern, which the 2025 update specifically scrutinizes
A burst of installs in one day registers as a velocity spike and gets discounted. A smooth, multi-day pattern gets credited. This is why drip-paced campaigns work and one-shot blasts don't; Keyword Installs Explained covers the mechanism in detail.
How conversion rate gates everything
Even with massive install velocity, if your store page converts at 1.5% when peers are at 5%, the store discounts your relevance score. The logic is straightforward: a low conversion rate means the store is wasting impressions on you, so it shows you less.
Conversion rate is driven by:
- Star rating (~20% lift per 0.5★, per Apple's published data)
- Review volume (1K+ reviews convert ~30% better than under 100)
- First screenshot (the only one most users see)
- App icon (smaller effect than the first two, but measurable)
- Subtitle / short description (smaller again, but measurable)
The first two are the highest-leverage. See The Hidden Math of App Ratings for the rating curve in detail.
What changed in 2026
Three notable algorithm updates:
Stricter velocity-spike detection (Apple, late 2025)
The algorithm now smooths your install velocity over a 7-day window. A single big day of installs gets averaged into the rest, which means the store doesn't see a spike, which means you don't get the rank lift you'd have gotten in 2023. Solution: drip across multiple days.
Apple Search Ads weighting bump (Apple, January 2026)
Apple slightly increased the weight of paid keyword bids on organic ranking for the same term. A team running an active Search Ads campaign on "habit tracker" gets a small organic boost for that term too. Effect size is modest (3-7% rank lift in our data) but real.
Review velocity smoothing (Google, March 2026)
Mirrors Apple's velocity smoothing but for reviews. A batch of 100+ reviews delivered in 48 hours gets weighted down. Spread the same batch across 14-21 days and the full impact lands.
What stays the same
Three things didn't change and are unlikely to:
- Metadata still matters. Title, subtitle (iOS), and keyword field (iOS) or short description (Android) anchor the relevance estimate the algorithm starts with. Skipping metadata optimization makes everything else harder.
- Engagement is still measured indirectly. Retention curves, session length, and uninstall rates feed into the long-term relevance estimate. ASO can move rank short-term; only a product people actually use sustains it.
- Competition still wins. If you're a 50-person team competing for "photo editor" against Adobe, no amount of ASO budget moves you above them. Pick a smaller pond first.
What this means for ASO strategy
If you understand the algorithm, your strategy more or less writes itself:
- Fix rating first, because it gates the conversion rate that gates everything else
- Run paced keyword install campaigns on terms where competition is winnable
- Maintain review velocity with a monthly small order, not quarterly large ones
- Pair keyword campaigns with Apple Search Ads on the same terms to stack the organic boost
- Watch rank at 24-hour granularity so you can react to algorithm pulses
The 5-strategy framework is laid out in 5 ASO Strategies That Doubled Our Customers' Installs. The data behind it is in We Analyzed 50,000 Apps.
Frequently asked questions
How does the App Store algorithm decide rankings?
It scores every candidate app against the user's query using eight signals: install velocity, conversion rate, star rating, review volume, metadata relevance, engagement, paid bids (Apple only), and competition. Install velocity is the heaviest weight you can influence.
Is the Google Play algorithm different from Apple's?
Mostly the same. Both weight install velocity, conversion rate, rating, and reviews heavily. The differences are in metadata parsing (Apple uses an explicit keyword field; Google parses the description) and in Apple's Search Ads integration which has no Google equivalent.
What changed in the 2026 ASO algorithm?
Three things: Apple smoothed install velocity to demote one-day spikes, Apple bumped the weight of Search Ads spend on organic rank for the same keyword, and Google smoothed review velocity to demote batched review drops.
How do I know if my rank is from the algorithm or just luck?
Charts. Pull daily rank snapshots over 30+ days and overlay them against your install events. If rank moves 3-7 days after install spikes and decays 14-30 days later, that's the algorithm. If rank wobbles randomly, that's noise.
Can I trick the App Store algorithm?
No. You can work with it, by moving the signals it weights, in patterns that look like real consumer demand. Tricking it (fake reviews, install fraud, keyword stuffing) gets detected within weeks and the resulting demotion is much harder to recover from than the original rank you were trying to fix.
For the practical playbook, see Step-by-Step: Launch a Keyword Install Campaign.
Put it into practice
See a live rank projection for your app, free. 50% off your first order.
Start free