BugChorus — App Review Analyzer & Release Monitor

Know what users are reporting after a mobile release without reading hundreds of reviews.

BugChorus turns public Apple App Store and Google Play reviews into ranked issue clusters, traceable review evidence, and a guarded version-over-version comparison. Use it for your own apps or any public competitor listing.

$0.44-$0.50 per non-empty app-country report · No Apple, Google, or LLM API keys · One-time analysis or scheduled monitoring

Run the preconfigured one-app example

What you learn from one run

Question	What BugChorus returns
What are users repeatedly complaining about?	Ranked crash, login, billing, performance, ads, lost-data, feature, update, and UX issue clusters.
Does the newest observed version look worse?	A guarded comparison against the prior observed version, including sample sizes and negative-review rates.
Which reviews support the signal?	Traceable evidence with rating, date, app version, excerpt, and source link when available.
What changed since the last check?	In monitor mode, the number of genuinely new reviews since the saved baseline.
Can I use the results elsewhere?	A readable HTML/Markdown report plus structured data exportable as JSON, CSV, or Excel.

The Actor returns ALERT, WATCH, CLEAR, or INSUFFICIENT_DATA. It uses INSUFFICIENT_DATA when there is not enough version-labeled review evidence for a responsible comparison.

Start in under a minute

1. Open the one-app example.
2. Replace the example with a public App Store or Google Play URL.
3. Choose the storefront country and review window.
4. Click Start and open HTML report when the run finishes.

The public example is configured for one iOS app in one country with a $0.51 maximum run charge.

Minimal input:

{
  "appUrls": [
    "https://apps.apple.com/us/app/instagram/id389801252"
  ],
  "countries": ["us"],
  "daysBack": 30,
  "mode": "snapshot"
}

Google Play URLs, ios:389801252, and android:com.example.app references are also accepted.

What the report looks like

Each report includes:

• app, platform, storefront country, store rating, and current observed version;
• reviews analyzed, issue-review count, and negative-review rate;
• top issue clusters with counts, share, severity, and version distribution;
• the strongest supporting review excerpts;
• current-versus-prior version sample sizes and rate difference;
• genuinely new reviews when stateful monitoring is enabled;
• direct links to the public store listing or review when available.

Why use this instead of a raw review export?

A raw export gives you rows to read. BugChorus adds the analysis layer needed for a release decision:

• a shared issue taxonomy across iOS and Android;
• ranked evidence instead of an unprioritized text dump;
• a minimum-sample guard before making version-regression claims;
• a human-readable decision report for product, QA, or release meetings;
• structured output for dashboards, webhooks, APIs, and automation;
• saved monitor state for identifying reviews that are actually new.

Classification is deterministic. There is no model prompt, LLM key, token bill, or changing AI interpretation between runs.

Best uses

• Release checks: investigate whether negative feedback increased after shipping a mobile update.
• Issue discovery: surface repeated crash, login, payment, performance, ad, or UX complaints.
• QA prioritization: attach review evidence to the issues most worth reproducing first.
• Competitor monitoring: watch public reviews for product gaps and recurring release problems.
• Regional comparison: analyze the same listing across up to five storefront countries.
• Scheduled monitoring: run a saved Apify task daily or weekly and route results into another workflow.

How the analysis works

1. The Actor fetches recent public reviews from the selected stores and countries.
2. It normalizes ratings, review text, dates, versions, helpful votes, and source links.
3. It groups selected low-rating reviews into practical issue families and ranks their evidence.
4. When version samples are strong enough, it compares the newest observed version with the prior observed version.

The regression calculation uses all fetched ratings. focusRatings controls which ratings become issue clusters and supporting evidence.

A newer version observed in review data can take precedence when store metadata is stale during a staged rollout. Both compared versions must meet minReviewsForRegression, and the rate change must pass minimum-difference and statistical-signal checks.

Input reference

Input	Purpose	Default
appUrls	Up to 20 App Store or Google Play URLs/IDs.	Required
countries	One to five two-letter storefront countries.	us
language	Google Play review language.	en
daysBack	Review age window from 1 to 365 days.	30
maxReviewsPerAppCountry	Maximum recent reviews fetched for each app-country target.	300
focusRatings	Ratings used for issue clusters and evidence.	1, 2, 3
minReviewsForRegression	Minimum reviews required for each compared version.	5
mode	One-time snapshot or stateful monitor.	snapshot
stateKey	Optional stable name for a monitor baseline.	Derived from apps
includeRawReviews	Include normalized review rows inside each report.	false
maxEvidencePerIssue	Maximum review examples retained per issue.	3

Output and integrations

Output	Use
Structured reports	One dataset item per app-country target; export as JSON, CSV, or Excel.
REPORT.html	Browser-ready report for a product, QA, or release meeting.
REPORT.md	Portable summary for GitHub, Slack, Notion, or an issue tracker.
OUTPUT	Machine-readable run totals, verdict counts, failures, and monitor details.

Save the input as an Apify task to schedule it. Use Apify's REST API, webhooks, integrations, or MCP support to send the normalized results into dashboards, alerts, or internal workflows.

Stateful monitoring

Set mode to monitor and use the same stateKey on later runs.

• The first run creates a baseline and reports zero new reviews.
• Later runs compare fetched review IDs with that baseline.
• Separate stateKey values keep products, clients, or environments isolated.
• A scheduled Apify task can run the same monitor daily or weekly.

Use a different stateKey whenever you want a completely separate monitor history.

Pricing

The app-report event is charged only after a successful app-country report containing at least one review is written.

Apify buyer tier	Price per report
Free / no discount	$0.50
Starter / Bronze	$0.48
Scale / Silver	$0.46
Business / Gold	$0.44

Examples at the highest per-report price:

• 1 app × 1 country = up to $0.50
• 2 apps × 2 countries = up to $2.00
• 10 apps × 1 country = up to $5.00

Invalid targets and valid targets with no reviews do not trigger the app-report charge. Apify may display its very small standard Actor-start event separately. Set a maximum run charge if you need a hard spending ceiling.

Limits and responsible interpretation

• Review signals are investigation evidence, not proof that a software release caused a defect.
• Confirm high-severity findings with telemetry, crash reporting, support tickets, and reproduction.
• Apple currently exposes at most 500 recent reviews per app and storefront through its public feed.
• The issue taxonomy is optimized for English text. Ratings and version statistics remain language-independent.
• Google Play fields and availability can vary by country, language, rollout, and upstream changes.
• The Actor reads public data only. It cannot access App Store Connect, Google Play Console, private analytics, crash logs, or customer records.
• It does not reply to reviews, modify store listings, or make release decisions automatically.

If one target fails, the run records a clear target-level error and continues with the remaining targets.

FAQ

Do I need Apple, Google, or LLM credentials?

No. BugChorus reads public store listings and reviews, and its classification is deterministic.

Can I analyze competitor apps?

Yes. You can analyze any public App Store or Google Play listing available in the selected storefront.

Why does the report say INSUFFICIENT_DATA?

There were not enough version-labeled reviews for both compared versions. Increase the review window, fetch more reviews, or cautiously lower minReviewsForRegression. Issue clusters still work without a version comparison.

Why does the first monitor run show zero new reviews?

The first run establishes the baseline. New-review counts become meaningful on the second run using the same stateKey.

Does ALERT prove the newest release is broken?

No. It means the public review evidence is strong enough to investigate. Confirm it using your internal engineering and support data.

Can I export raw reviews?

Yes. Enable includeRawReviews. Leave it off when issue clusters and evidence excerpts are sufficient and you want smaller reports.

Support

When reporting a problem, include the Apify run URL, public app URL, storefront country, and input settings. Never include account passwords, store-console credentials, private API keys, or customer data.