# App Store & Google Play Reviews + AI Classifier | $0.10/1K (`typedrows/app-reviews-classifier`) Actor

$0.10 per 1,000 reviews, +$0.50 per 1,000 classified. Scrape App Store and Google Play reviews, then classify: main issue (crash, login, payment, ads, support, UX), severity score, churn flag, each with a confidence number. Complaint and sentiment analysis at one tenth of LLM cost. CSV or JSON.

- **URL**: https://apify.com/typedrows/app-reviews-classifier.md
- **Developed by:** [Typedrows](https://apify.com/typedrows) (community)
- **Stats:** 2 total users, 1 monthly users, 100.0% runs succeeded, 0 bookmarks
- **User rating**: No ratings yet

## Pricing

from $0.50 / 1,000 review classifieds

This Actor is paid per event. You are not charged for the Apify platform usage, but only a fixed price for specific events.

Learn more: https://docs.apify.com/actors/running/actors-in-store.md#pay-per-event

## What's an Apify Actor?

An Actor is a serverless cloud program that runs on the Apify platform. It has two run modes.
In Batch mode, an Actor accepts a well-defined JSON input, performs an action which can take anything from a few seconds to a few hours,
and optionally produces a well-defined JSON output, datasets with results, or files in key-value store.
In Standby mode, an Actor provides a web server which can be used as a website, API, or an MCP server.

Apify vocabulary and the platform model are defined once, in the agent quickstart at https://apify.com/agents.md.

## How to integrate an Actor?

If asked about integration, you help developers integrate Actors into their projects.
You adapt to their stack and deliver integrations that are safe, well-documented, and production-ready.

Do not guess an integration path. Every one of them is in the agent quickstart at https://apify.com/agents.md: the Apify MCP server, Agent Skills with the Apify CLI, the JavaScript and Python clients, the REST API, and the account-free path for an agent with no human to sign in. It also carries the rule on stating cost before the first paid run.

For examples already wired to this Actor's own input schema, see the [API](#api) section below.

Each client library has reference documentation the quickstart does not restate: [JavaScript/TypeScript](https://docs.apify.com/api/client/js/docs.md) (`npm install apify-client`) and [Python](https://docs.apify.com/api/client/python/docs.md) (`pip install apify-client`).

# README

## App Reviews Classifier

Scrapes App Store and Google Play reviews for any list of apps and, by default, adds **typed classification columns** to every review using [jev](https://typesafe.ai), a non-generative classifier. You get a dataset you can pivot on immediately: main issue, severity, and boolean-style flags — no prompt engineering, no LLM bill.

Classification columns cost about one tenth of LLM-based alternatives.

### Input

| Field | Type | Default | Description |
|---|---|---|---|
| `apps` | array of strings | *required* | App Store numeric ids or URLs (`310633997`, `https://apps.apple.com/us/app/whatsapp-messenger/id310633997`) and Google Play package ids or URLs (`com.whatsapp`, `https://play.google.com/store/apps/details?id=com.whatsapp`). All digits or an `apps.apple.com` URL means App Store, anything else Google Play. |
| `countries` | array of strings | `["us"]` | Two-letter store country codes. Every app is scraped in every country. |
| `language` | string | `"en"` | Google Play review language (ignored for the App Store). |
| `maxReviewsPerApp` | integer | `500` | Per app and country, 1–20000. |
| `classify` | boolean | `true` | Add the classification columns. |
| `questions` | object | preset | Custom jev question schema replacing the preset (see below). |
| `confidenceThreshold` | number | `0.8` | Rows with any choice/score confidence below this get `review_needed: true`. |

### Output

One dataset item per review.

| Field | Description |
|---|---|
| `platform` | `"appstore"` or `"googleplay"` |
| `appId` | Numeric App Store id or Google Play package id |
| `reviewId` | Store review id |
| `date` | ISO 8601 |
| `rating` | 1–5 |
| `title` | App Store only, `null` on Google Play |
| `text` | Review body |
| `author` | Reviewer name |
| `version` | App version reviewed |
| `country` | Store country code |
| `thumbsUp` | Google Play only, `null` on the App Store |
| `replyText` | Developer reply, Google Play only |
| `url` | Link back to the review / reviews page |

When `classify` is on, each question adds a column named after the question:

| Question | Type | Column |
|---|---|---|
| `issue` | choice | `issue` (label) + `issue_confidence` |
| `severity` | score | `severity` (0–3) + `severity_confidence` |
| `mentions_crash` | noul | `mentions_crash` (probability 0–1) |
| `mentions_payment` | noul | `mentions_payment` |
| `mentions_support` | noul | `mentions_support` |
| `mentions_update` | noul | `mentions_update` |
| `would_churn` | noul | `would_churn` |

Plus `review_needed` (true when any choice/score confidence is under `confidenceThreshold`). `issue` labels: `crash_bug`, `login_account`, `payment_billing`, `performance`, `ads`, `support`, `ux_design`, `feature_request`, `content_quality`, `praise`, `other`.

Reviews with empty text get `null` classification columns and are not charged. A review whose classification failed after retries gets a `classification_error` string and is not charged either.

#### Custom questions

Replace the preset with up to 50 of your own:

```json
{
  "is_refund_request": { "type": "noul", "instructions": "the reviewer asks for a refund" },
  "topic": { "type": "choice", "instructions": "what the review is about",
             "criteria": { "price": "cost or value for money", "speed": "how fast the app is" } },
  "urgency": { "type": "score", "instructions": "how urgent is this",
               "criteria": ["not urgent", "somewhat urgent", "drop everything"] }
}
```

`choice` criteria is an object of `{label: description}`, `score` criteria is a **list** of descriptions from low end to high end. Bad schemas fail before anything is scraped.

### Pricing (pay per event)

| Event | Charged |
|---|---|
| Review scraped (built-in dataset item) | Once per review row written, USD 0.0001 |
| Review classified | Once per review successfully classified, USD 0.0005 |

Set `classify: false` to pay only for scraping.

### Requirements

Set `TYPESAFE_API_KEY` as a secret environment variable on the actor (only needed when `classify` is true). Get one at [typesafe.ai](https://typesafe.ai).

### Limitations

- **App Store: 500 newest reviews per country, maximum.** That is the ceiling of Apple's public RSS feed, not of this actor. Add more `countries` to get more reviews; there is no way to reach older history through this endpoint.
- App Store reviews have no thumbs-up count or developer reply.
- Google Play returns reviews in one language/country at a time; both are inputs.
- jev sees the first 2,000 characters of a review.

### Run it locally

```bash
npm install
node test.js                                    # self-check, hits all three live APIs
APIFY_LOCAL_STORAGE_DIR=./storage node --env-file=.env src/main.js
```

with `storage/key_value_stores/default/INPUT.json` holding the input and `.env` holding `TYPESAFE_API_KEY=...`.

# Actor input Schema

## `apps` (type: `array`):

App Store numeric ids or URLs, and Google Play package ids or URLs. All digits or an apps.apple.com URL is treated as App Store; anything else as Google Play.

## `countries` (type: `array`):

Two-letter store country codes.

## `language` (type: `string`):

Google Play review language (ignored for the App Store).

## `maxReviewsPerApp` (type: `integer`):

Per app and country. The App Store RSS feed caps at 500 newest reviews per country.

## `classify` (type: `boolean`):

Run every review through jev and add the typed columns.

## `questions` (type: `object`):

Optional jev question schema replacing the preset. Up to 50 questions of type noul, choice (criteria = {label: description}) or score (criteria = list of descriptions, low end first).

## `confidenceThreshold` (type: `number`):

Rows with any choice/score confidence below this get review\_needed: true.

## Actor input object example

```json
{
  "apps": [
    "https://apps.apple.com/us/app/whatsapp-messenger/id310633997",
    "com.whatsapp"
  ],
  "countries": [
    "us"
  ],
  "language": "en",
  "maxReviewsPerApp": 500,
  "classify": true,
  "confidenceThreshold": 0.8
}
```

# Actor output Schema

## `reviews` (type: `string`):

One item per review: raw store fields plus issue, severity, flags and confidence columns.

# API

You can run this Actor programmatically using our API. Below are code examples in JavaScript, Python, and CLI, as well as the OpenAPI specification and MCP server setup.

## JavaScript example

```javascript
import { ApifyClient } from 'apify-client';

// Initialize the ApifyClient with your Apify API token
// Replace the '<YOUR_API_TOKEN>' with your token
const client = new ApifyClient({
    token: '<YOUR_API_TOKEN>',
});

// Prepare Actor input
const input = {
    "apps": [
        "https://apps.apple.com/us/app/whatsapp-messenger/id310633997",
        "com.whatsapp"
    ]
};

// Run the Actor and wait for it to finish
const run = await client.actor("typedrows/app-reviews-classifier").call(input);

// Fetch and print Actor results from the run's dataset (if any)
console.log('Results from dataset');
console.log(`💾 Check your data here: https://console.apify.com/storage/datasets/${run.defaultDatasetId}`);
const { items } = await client.dataset(run.defaultDatasetId).listItems();
items.forEach((item) => {
    console.dir(item);
});

// 📚 Want to learn more 📖? Go to → https://docs.apify.com/api/client/js/docs

```

## Python example

```python
from apify_client import ApifyClient

# Initialize the ApifyClient with your Apify API token
# Replace '<YOUR_API_TOKEN>' with your token.
client = ApifyClient("<YOUR_API_TOKEN>")

# Prepare the Actor input
run_input = { "apps": [
        "https://apps.apple.com/us/app/whatsapp-messenger/id310633997",
        "com.whatsapp",
    ] }

# Run the Actor and wait for it to finish
run = client.actor("typedrows/app-reviews-classifier").call(run_input=run_input)

# Fetch and print Actor results from the run's dataset (if there are any)
print(f"💾 Check your data here: https://console.apify.com/storage/datasets/{run.default_dataset_id}")
for item in client.dataset(run.default_dataset_id).iterate_items():
    print(item)

# 📚 Want to learn more 📖? Go to → https://docs.apify.com/api/client/python/docs/quick-start

```

## CLI example

```bash
echo '{
  "apps": [
    "https://apps.apple.com/us/app/whatsapp-messenger/id310633997",
    "com.whatsapp"
  ]
}' |
apify call typedrows/app-reviews-classifier --silent --output-dataset

```

## MCP server setup

```json
{
    "mcpServers": {
        "apify": {
            "type": "http",
            "url": "https://mcp.apify.com/?tools=fetch-actor-details,typedrows/app-reviews-classifier"
        }
    }
}
```

The hosted server signs you in with OAuth on first connect, so no API token belongs in this config. Clients without OAuth support can send an `Authorization: Bearer <APIFY_API_TOKEN>` header instead, using a token from API & Integrations in Apify Console (https://console.apify.com/settings/integrations).

## OpenAPI specification

Download the OpenAPI definition: https://api.apify.com/v2/actors/AiWePVJAIeYijLmjB/builds/mT5JayVx3D4nhArQ5/openapi.json