When an app is available on iOS and Android, I want to read the feedback from both stores without switching between pages. Are people raising the same issues, or does one platform need a closer look?
I built CleanScrape's App Store and Google Play Reviews Scraper to put those reviews into one table. This walkthrough uses my paid Actor to collect a small sample, then shows how to compare the comments in a spreadsheet. The comparison is something you do after export, not an automatic analysis feature.
Let's use Spotify as an example. You can replace its store links with those for your own app or another publicly listed app.
Watch the export workflow
The short demo shows the app links, regional settings, review table and spreadsheet export. It requests up to 20 reviews per store. The comparison steps below pick up where the video leaves off.
Get your first small export
Open the ready-made Spotify example. Its preset is 100 reviews per store; change Maximum reviews to 20 to follow the video.
- Keep both stores selected. For another app, paste its Google Play and Apple App Store links into the corresponding fields.
- Choose the country and Google Play language by name. The country selects a storefront, not the reviewer's verified location; the language setting does not translate comments.
- Use newest reviews and keep the 1 GB memory setting. Set a spending limit before starting.
- After the run, open Review essentials for the compact table. Open Run summary and field guide to check that both requested stores returned data.
- Export CSV or Excel. Keep the source column in your spreadsheet.
The maximum is per store: 20 plus 20 means up to 40 reviews. Availability can produce fewer. If Apple supplies an unverified empty feed, the Actor now reports a source error rather than implying that the app has no reviews. A Google Play export can still succeed in that run, so read the per-store report before comparing platforms.
For an API workflow, the same small request is:
{
"store": "both",
"googlePlayAppId": "https://play.google.com/store/apps/details?id=com.spotify.music",
"appStoreId": "https://apps.apple.com/us/app/spotify/id324684580",
"country": "us",
"language": "en",
"maxReviews": 20,
"sort": "newest"
}
Existing package names and numeric app IDs still work.
Compare the feedback, not just the averages
Start with a specific question, such as: Are the low-rated reviews describing similar problems on iOS and Android?
- Filter rating to 1 or 2 and read the review text. Keep source, date and appVersion visible.
- Add your own issue column. Use labels that fit the comments you actually read, such as sign-in, playback or billing. Those are possible labels, not findings from this sample.
- Group or make a pivot table by issue and source, counting review IDs. This gives you a reading list of recurring themes in the downloaded sample.
- Go back to the original comments before drawing a conclusion. A short negative review may not explain its cause. If appVersion is available, compare versions within each store, not as interchangeable iOS and Android release numbers.
This is not sentiment analysis, a satisfaction score or proof that one release caused a complaint. Twenty recent reviews per store can help you decide what to investigate; it cannot establish how common an issue is among all users. The two exports may also cover different date spans.
Know what the columns mean
| Field | How I would use it |
|---|---|
| source | Keep iOS and Android feedback separate while using the same spreadsheet. |
| rating and text | Find comments to read, then understand the actual feedback. |
| date | Check the source timestamp; Apple's value is an update time, not necessarily the original publication date. |
| appVersion | Inspect version-specific feedback when the source supplies it. Missing versions stay empty. |
| reviewId | Identify reviews when joining or updating exports. |
| developerReply | Read available Google Play replies. Apple replies are not supplied by this feed. |
Google Play titles are also empty in this adapter. Missing source fields stay empty rather than being filled with guesses. Choose the full Reviews view if you need every output field.
What the whole run costs
At the current base rate and 1 GB memory:
| Delivered output | Result charges | Startup | Total |
|---|---|---|---|
| 40 reviews | $0.008 | $0.01 | $0.018 |
| 200 reviews | $0.04 | $0.01 | $0.05 |
The second row is the existing preset if it returns 100 reviews from each store. Startup is charged even when no reviews return. Selecting more memory increases the startup charge; saved tasks can retain their own memory setting. Filtering rows in Excel does not change the number already delivered and billed.
These are base prices before subscription discounts, checked on 27 September 2026. See the Actor's Pricing tab for your applicable rate.
If you repeat the export
Apple's public feed is a recent, limited source, not a complete archive. Source availability can change. Increasing the maximum does not guarantee more history.
The Actor does not maintain review history across separate runs. Where a review ID exists, use the combination of source, appId, country and reviewId as a starting key. Update edited reviews instead of appending duplicates, and handle missing IDs separately.
Open the example, change the app links and keep the first run small. No store login or personal cookies are needed. CleanScrape is independent of Apple, Google and Spotify.
If you use review exports, what makes them useful after the download: comparing platforms, reviewing a release, or feeding another workflow? Practical feedback helps me choose what to improve. For a bug, open an Actor issue with the run ID and a non-sensitive input example. Keep tokens and credentials out of comments.
Use the same export in an assistant
I also published a free app-review skill for assistants that support agent skills. It gives the assistant a way to compare the two stores without losing the review IDs or mistaking a small sample for every user's opinion.
Use these saved review exports only. Compare iOS and Android complaints, show the rating counts for each store, and flag missing data. Do not start a new run.
You can also supply two snapshots to find newly observed or edited reviews. This is analysis of your files, not an automatic monitoring service. The instructions are free; fresh Actor runs and your assistant's own usage are separate.
Top comments (0)