Google Play Store Scraper - Apps & Ratings avatar

Google Play Store Scraper - Apps & Ratings

Pricing

Pay per event

Go to Apify Store
Google Play Store Scraper - Apps & Ratings

Google Play Store Scraper - Apps & Ratings

Read Google Play apps by package name or search: title, developer, category, rating, install band, description and icon. Returns all three of Play's differing review counts under names that say what they are.

Pricing

Pay per event

Rating

0.0

(0)

Developer

SR

SR

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

5 days ago

Last modified

Categories

Share

Read apps from the Google Play store as structured data: title, developer, category, rating, install band, description and icon.

Search Play, or give package names directly. No API key, no browser.

Three review counts, because Play publishes three

This is the thing to know before you compare two apps on "reviews".

For Spotify, Play's page displays 25,202,278 reviews. Its own data carries 36,266,585 and 27,087,213 as well. Three numbers, all real, all different:

fieldSpotify
reviews_shown25,202,278
ratings_total36,266,585
rating_histogram_total27,087,213

They count different things: written reviews, every rating including the silent ones, and the total behind the star breakdown. Pick one and call it "reviews" and any comparison between two apps quietly stops meaning anything, because you have no way of knowing whether the other app's number came from the same place.

All three are returned, named for what they are. Use reviews_shown to quote what a visitor sees, and ratings_total when you need a number you can sum.

The developer field that is not the developer

Play carries a field that reads like the developer and is the merchant of record. In the United States it happens to hold the developer's name, so it looks correct. In the United Kingdom it reads "Google Commerce Ltd" for every single app.

A column that is right in one market and uniformly wrong in every other is worse than no column, because nothing about it looks broken. So developer is the actual developer (Spotify AB, Calm.com Inc.) in every market, and seller holds the merchant of record separately.

Search or package names

Search uses Play's own store search, so the order you get is Play's ranking. vpn returned Proton VPN, Secure VPN, Turbo VPN and the rest in the order Play put them.

Package names (com.spotify.music) or Play links skip the search and read those apps directly. Use both together and the named apps come first.

Play is a per-country store, and this says which one it read

Availability, pricing, ranking and, as above, even the developer field differ by market. Country storefront sets both the storefront and where the request appears to come from, and market is recorded on every row, so two runs are comparable rather than accidentally different.

What you get per app

  • title, developer, seller, genre
  • rating out of 5, plus the three counts above
  • installs_text (1,000,000,000+) and installs as a number
  • description, icon_url, updated_at as a UTC timestamp
  • app_id, url, market

When Google moves something, you will know

Play does not label its data. Every field is addressed by its position in nested arrays, which means a path can start returning the wrong thing, or nothing, without any error at any point. That is how scrapers of this store quietly rot.

Two things guard against it. Every field is checked against what it should look like before it is published: a rating has to be a number between 0 and 5, so a path that starts returning a developer name is dropped rather than published. And the run summary reports fieldsThatDidNotResolve with a count per field.

A healthy run reports nothing there. A run reporting {"updated_at": 3} tells you three apps did not publish an update date, which is normal. A run reporting most fields at once tells you Google changed the page and the paths need revisiting, on the day it happens rather than months later.

Scale

Each app is one request, and a search is one more. Four apps took ten seconds and eight search results took seventeen. Maximum apps caps both time and cost.

What people use this for

App store optimisation. Rank across a search term, with rating and install band beside each result, is the whole competitive picture for a keyword.

Competitor tracking. Run a set of package names on a schedule and keep the rows. Rating and install band move slowly, which is exactly why the movement is worth catching.

Market comparison. The same app across several country storefronts, with market on every row, shows where it is promoted and where it is not.

Portfolio monitoring. Your own apps and your rivals' in one table, refreshed whenever you like, at one request each.

Notes

Only public store listings are readable. An app that is unpublished or not available in the chosen country is reported by name rather than returned as an empty row.

Ratings and install bands move, so a run is a snapshot. updated_at is the app's own last update, not the time of the run.

Reviews are counted but not fetched: the counts are here, the review texts are a separate surface and are not part of this Actor.