App Ratings and Reviews: How They Work and Why They Matter

Star ratings are the first thing a person sees next to an app's name, and the stores use them as a quality signal long before anyone reads a review. This guide explains how ratings are collected and counted on the App Store and Google Play, what the rules allow, how both companies police fakes, and why the number matters for charts.

What a rating is, and where it comes from

On both stores a rating is a one-to-five star score, and a review is optional text attached to it. Apple builds a summary rating from individual ratings and shows it on the product page and in search results; that summary is calculated separately for each storefront territory, so an app can be 4.7 in Germany and 4.2 in Brazil at the same time ¹.

Google Play goes a step further. Since November 2021 the rating shown to phone users is specific to the country tied to their Play account, and from 2022 it also varies by device type (phone, tablet, Chromebook, Wear and so on). Google's stated reason was that a bug affecting one region or form factor should not drag down the score everywhere ⁶. On top of that, the public Play rating is weighted toward recent ratings rather than being a lifetime average; the lifetime figure is shown separately in Play Console ⁷.

Apple does not weight by recency. Instead it gives developers a manual lever: you can reset the summary rating when you ship a new version. Apple suggests doing this sparingly, because a low rating count can put people off, and the reset does not remove written reviews, which keep showing on the product page ¹.

Asking for a rating: the built-in prompts

Both platforms provide a system prompt that lets a user rate without leaving the app, and both expect you to use it.

Apple's guidance is to ask after someone finishes a task, level or purchase rather than mid-flow, and to keep support contact details easy to find ¹. Guideline 5.6.1 requires the Apple-provided API for review prompts and says custom prompts will be rejected; guideline 3.2.2 adds that an app must not gate features or content behind rating it ².

Google's documentation warns against wiring the review flow to a button, because a user who has hit the quota would tap it and see nothing. If you need a "Rate us" link in settings, send people to the store listing instead ³. During testing, the quota is not enforced for builds installed from the internal test track, and builds shared through internal app sharing cannot submit reviews at all ⁴.

What the policies prohibit

The two stores phrase it differently, but the core rule is identical: ratings must be earned, not bought.

Google Play's User Ratings, Reviews and Installs policy bans inflating ratings, reviews or install counts by illegitimate means. Its worked examples include offering a discount in exchange for a high rating and repeatedly submitting ratings while posing as users. It also says replies to reviews should not ask for a higher rating ⁵. Google has described a combined human-plus-machine-learning system for catching fake and incentivised ratings, including coordinated five-star campaigns and one-star attacks ¹⁰.

Apple's guideline 5.6.3, "Discovery Fraud", says manipulating any element of the customer experience, including charts, search and reviews, is not permitted, and the Business section states that paid, incentivised, filtered or fake feedback can lead to expulsion from the Developer Program ².

Outside the stores, the US Federal Trade Commission finalised a rule in August 2024 (effective October 2024) that bans buying or selling fake reviews and allows civil penalties against knowing violators ¹²; legal commentaries put the 2025 per-violation ceiling in the low $50,000s unconfirmed.

Enforcement in numbers

Apple publishes an annual fraud report. The relevant figures for ratings and reviews:

Roughly one in seven submitted ratings or reviews was rejected in 2025 ⁸. Google's 2025 safety report (February 2026) counts more than 1.75 million policy-violating apps stopped and over 80,000 developer accounts banned, but does not break out review fraud ¹¹; its last public description of review removals spoke of "millions" per week ¹⁰.

Benchmarks: what counts as a good rating

Neither Apple nor Google publishes average ratings by category. The closest public benchmark comes from AppTweak's analysis of the top 1,000 US apps and games across full-year 2025: around 95% of apps featured on the App Store were rated 4.0 or higher and 65% were at 4.6 or higher, while roughly 96% of featured Google Play apps were at 4.0 or above. Fewer than 5% of featured apps on either store sat below four stars ¹³.

The same dataset found that 4-star apps ranked in the top 10 for an average of 4.4 keywords, against 2.2 for 3-star apps and 1.3 for 1-star apps. AppTweak itself flags this as correlation, not causation, since better-rated apps also tend to be larger and older ¹³.

Practical reading: 4.0 is the floor below which featuring becomes unlikely, 4.5 or higher is where most featured apps live, and volume matters, since a 4.8 from 30 people is less convincing than a 4.6 from 30,000 ¹.

How ratings affect conversion and visibility

Neither store documents the weight of ratings in search ranking or featuring, and the only honest answer is "it is a factor". Google says plainly that ratings are a download signal and feed into featuring and placement ⁶. Apple's guideline 5.6.4 lists excessive negative reviews among the indicators of poor app quality that can count against a developer ².

Conversion is where the effect is easiest to see: the star appears in search results and above the fold on every product page, so a drop from 4.6 to 3.9 is visible before a user reads a word. Apple now also generates review summaries with a large language model and refreshes them continuously; the feature launched in the US and, as of October 2026, has rolled out to eight more English-language storefronts ¹. A cluster of complaints about a bug can therefore surface in a one-paragraph summary even if the overall score stays high.

How this shows up in the charts

Top charts rank by downloads and revenue, not by stars, so a rating does not move an app up a chart directly. It moves the inputs. A higher rating lifts the share of product-page visitors who install, and installs are what the free charts count. On Google Play the recency weighting means a buggy release can drop an app's displayed rating within weeks, which shows up as a conversion dip and then a chart slide, while the same app on iOS keeps its lifetime score unless the developer resets it ^7,1.

When an app jumps the charts on AppWinners and its rating count climbs in step, that is usually a good release meeting a sensible prompt strategy. When the rating count outruns the download rank, remember that Apple removed nearly 195 million ratings and reviews last year and blocked 11,500 apps from the charts ⁸.