iRecommend Product Reviews Scraper

Collect public product and service reviews from iRecommend.ru as clean, structured data.

Turn review cards into rows with stable IDs, ratings, authors, publication dates, comments, images, and direct links.

Use the data in spreadsheets, dashboards, warehouses, sentiment pipelines, or scheduled reputation monitoring.

• 🇷🇺 Preserves Russian and Cyrillic text exactly
• ⭐ Supports rating filters
• 📅 Supports publication-date filters
• 🔗 Accepts multiple product pages
• 📦 Exports JSON, CSV, Excel, XML, and RSS through Apify
• 🔄 Paginates automatically until your limit is reached

What does iRecommend Product Reviews Scraper do?

The Actor opens public irecommend.ru/content/... product or service pages and extracts one dataset row per review card.

It handles iRecommend's anonymous browser verification flow, follows pagination, and stops at your global review limit.

Every saved row includes:

• the source product page
• stable review, product, and version identifiers
• the direct full-review URL
• title and teaser text
• star rating
• author name, profile, and avatar
• publication date
• comment count
• preview image URLs and total photo count
• an ISO scrape timestamp

No iRecommend account is required.

Who is it for?

Brand and reputation teams

Monitor fresh feedback about a brand, delivery service, cosmetic, appliance, app, or other reviewed product.

Ecommerce analysts

Compare rating distributions, review cadence, engagement, and visual-review volume across competing products.

Consumer-insights researchers

Build a Russian-language corpus for topic analysis, complaint clustering, qualitative research, or sentiment scoring.

Data and automation teams

Schedule recurring runs and send normalized review records into a warehouse, BI tool, webhook, or downstream Actor.

Agencies

Create repeatable review-monitoring workflows for multiple clients without copying cards manually.

Why use this Actor?

Manual review collection is slow and difficult to refresh.

This Actor provides:

• predictable structured fields
• stable numeric IDs for deduplication
• automatic pagination
• multi-page input in one run
• filters that reduce downstream cleanup
• incremental dataset writes
• partial results if a later target page stops responding
• standard Apify exports and integrations

The default input is intentionally small, so you can validate the output before increasing maxItems.

What iRecommend review data can you extract?

Field	Type	Meaning
sourceUrl	string	Canonical product or service page
reviewId	integer	Stable iRecommend review ID
productId	integer	Product entity ID
versionId	integer	Review version ID exposed by the card
reviewUrl	string	Direct public review URL
title	string	Review headline
teaser	string	Review preview text
rating	integer	Star rating from 1 to 5
author	object	Name, profile URL, and optional avatar
publishedAt	string	Source date in DD.MM.YYYY format
commentsCount	integer	Visible comment count
imageUrls	array	Visible review preview images
photosCount	integer	Total photos reported on the card
scrapedAt	string	UTC extraction timestamp

How to scrape iRecommend product reviews

1. Open an iRecommend product or service page.
2. Copy its https://irecommend.ru/content/... URL.
3. Open this Actor's Input tab.
4. Add the URL under Product review pages.
5. Keep the default maximum for a cheap first run.
6. Optionally choose rating or date filters.
7. Click Start.
8. Open the dataset when the run finishes.
9. Export the records in your preferred format.

Add several product pages when you want one combined dataset.

Input

Example:

{
  "startUrls": [
    {
      "url": "https://irecommend.ru/content/sait-prilozhenie-dostavki-produktov-samokat"
    }
  ],
  "maxItems": 50,
  "minRating": 1,
  "maxRating": 3,
  "dateFrom": "2026-01-01",
  "dateTo": "2026-12-31"
}

startUrls

Required array of public iRecommend product or service review pages.

The Actor accepts request-list objects and plain URL strings through the API.

maxItems

Maximum reviews saved across all supplied pages.

The default is 10 and the maximum is 10,000.

minRating and maxRating

Optional integer boundaries from 1 through 5.

Use maxRating: 2 to focus on critical feedback.

Use minRating: 5 to collect only top-rated reviews.

dateFrom and dateTo

Optional inclusive dates in YYYY-MM-DD format.

The source's DD.MM.YYYY dates are parsed before filtering.

proxyConfiguration

Optional Apify Proxy or custom proxy settings.

Direct anonymous browser access is the default and normally completes iRecommend's public verification flow automatically.

Output example

{
  "sourceUrl": "https://irecommend.ru/content/sait-prilozhenie-dostavki-produktov-samokat",
  "reviewId": 11205258,
  "productId": 5814124,
  "versionId": 19755845,
  "reviewUrl": "https://irecommend.ru/content/samokat-sponsor-moikh-uzhinov-s-indeikoi-i-urodlivogo-kulicha-na-paskhu-dorogo-ili-vygodno",
  "title": "Самокат - спонсор моих ужинов...",
  "teaser": "Я очень люблю вкусную и полезную пищу...",
  "rating": 4,
  "author": {
    "name": "Panna Lisa",
    "url": "https://irecommend.ru/users/panna-lisa",
    "avatar": "https://cdn-irec.r-99.com/sites/default/files/imagecache/60i/pictures/picture-582516-e513c.jpg"
  },
  "publishedAt": "14.07.2026",
  "commentsCount": 12,
  "imageUrls": [
    "https://cdn-irec.r-99.com/sites/default/files/imagecache/200i/user-images/582516/example.jpg"
  ],
  "photosCount": 30,
  "scrapedAt": "2026-07-15T08:00:00.000Z"
}

How pagination and limits work

The first product page contains the first review batch.

Additional batches use iRecommend's ordinary ?page=N pagination.

The Actor reuses one browser session for each source page and stops when:

• maxItems is reached
• no next page exists
• a page produces no new review IDs
• the target fails after bounded waiting

Rating and date filters are applied before records are saved.

How deduplication works

Review rows are deduplicated by reviewId across the complete run.

This avoids duplicate exports when:

• the same product URL is supplied twice
• cards overlap during pagination
• multiple source pages expose the same review

For monitoring across separate runs, deduplicate downstream on reviewId or use it as an upsert key.

How much does it cost to scrape iRecommend reviews?

This Actor uses pay-per-event pricing.

You pay a small run-start fee and a per-review fee only for records saved to the dataset.

Filtered-out cards are not charged as review results.

Apify subscription tiers receive tiered per-result prices; check the live pricing panel for the exact current amount.

A small 10-review test is the best way to estimate the cost of your workflow.

Review monitoring workflow

For recurring reputation monitoring:

1. Create a task from a working Actor input.
2. Set a suitable maxItems limit.
3. Use dateFrom for the monitoring window.
4. Schedule the task daily or weekly.
5. Connect the dataset to your destination.
6. Upsert records by reviewId.
7. Alert on low ratings or selected Russian keywords.

Remember that a date filter does not change the source's page order; the Actor still reads public cards until it reaches the limit or the end.

Tips for reliable runs

• Start with one known public product page.
• Use canonical /content/... URLs rather than search pages.
• Keep first runs small.
• Increase limits only after inspecting sample records.
• Use multiple source URLs for a combined portfolio export.
• Use stable IDs instead of titles for deduplication.
• Do not block scripts in custom browser infrastructure; the anonymous verification needs JavaScript.
• Add a proxy only when direct routing is unavailable in your environment.

Integrations

Google Sheets

Send dataset rows to a spreadsheet for triage, tagging, and analyst notes.

Slack or email alerts

Filter for one- or two-star records and notify a customer-support or reputation channel.

Webhooks

Trigger downstream processing when a run succeeds and fetch the dataset through the API.

Data warehouses

Load JSON or CSV into BigQuery, Snowflake, PostgreSQL, or another analytics store.

Sentiment analysis

Pass title and teaser to a Russian-language sentiment or topic model.

Zapier and Make

Use Apify integrations to move new review records into no-code workflows.

API usage in JavaScript

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('automation-lab/irecommend-product-reviews-scraper').call({
  startUrls: [{
    url: 'https://irecommend.ru/content/sait-prilozhenie-dostavki-produktov-samokat',
  }],
  maxItems: 25,
  maxRating: 3,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Use with the Apify API in Python

from apify_client import ApifyClient

client = ApifyClient('YOUR_APIFY_TOKEN')
run = client.actor('automation-lab/irecommend-product-reviews-scraper').call(run_input={
    'startUrls': [{
        'url': 'https://irecommend.ru/content/sait-prilozhenie-dostavki-produktov-samokat'
    }],
    'maxItems': 25,
    'minRating': 4,
})

items = client.dataset(run['defaultDatasetId']).list_items().items
print(items)

Use with cURL

curl -X POST \
  'https://api.apify.com/v2/acts/automation-lab~irecommend-product-reviews-scraper/runs?token=YOUR_APIFY_TOKEN' \
  -H 'Content-Type: application/json' \
  -d '{
    "startUrls": [{
      "url": "https://irecommend.ru/content/sait-prilozhenie-dostavki-produktov-samokat"
    }],
    "maxItems": 10
  }'

Poll the returned run ID, then read its default dataset.

Use iRecommend reviews with Apify MCP

Connect AI assistants to this Actor through Apify MCP:

https://mcp.apify.com/?tools=automation-lab/irecommend-product-reviews-scraper

Claude Code setup

Add the Apify MCP server from your terminal:

$claude mcp add --transport http apify 'https://mcp.apify.com/?tools=automation-lab/irecommend-product-reviews-scraper'

Authenticate with your Apify token when prompted.

Claude Desktop setup

Add this server to the mcpServers object in Claude Desktop's configuration:

{
  "mcpServers": {
    "apify-irecommend": {
      "url": "https://mcp.apify.com/?tools=automation-lab/irecommend-product-reviews-scraper"
    }
  }
}

Cursor setup

Open Cursor Settings → MCP, add a remote HTTP server, and use the same Apify MCP URL. The mcpServers JSON above can also be adapted to Cursor's MCP configuration.

VS Code setup

Open MCP: Add Server from the VS Code command palette, choose HTTP, and paste the Actor-specific Apify MCP URL.

Example prompts

• “Collect the newest 25 reviews from this iRecommend product page.”
• “Find one- and two-star iRecommend reviews and summarize recurring complaints.”
• “Compare rating and comment patterns across these three iRecommend pages.”
• “Export review IDs, authors, ratings, comments, and direct URLs as a CSV-ready table.”

The Actor's structured dataset lets assistants reason over IDs, ratings, dates, and authors without parsing raw HTML.

Data quality and source behavior

The Actor extracts values shown on public review cards.

teaser is a preview, not necessarily the complete review body.

imageUrls contains visible preview images, while photosCount can be larger.

Comment totals and cards can change between runs as iRecommend updates the site.

The source date remains in DD.MM.YYYY format to preserve the displayed value.

Error handling

An invalid non-iRecommend URL is skipped with a warning.

If one source page fails, records already saved remain available.

If a later pagination page fails, the Actor keeps partial progress and logs the failing URL.

The run fails early when:

• no product URLs are supplied
• maxItems is outside the allowed range
• minimum rating is greater than maximum rating

Legality: is it legal to scrape iRecommend reviews?

This Actor is designed for publicly accessible review cards.

Data-protection, copyright, database-right, and website terms can vary by jurisdiction and use case.

You are responsible for ensuring that your collection, storage, and use have a lawful basis.

Avoid collecting unnecessary personal data, respect applicable retention rules, and do not use the output for harassment, spam, or prohibited profiling.

When in doubt, consult qualified legal counsel.

FAQ: frequently asked questions

Does the Actor require an iRecommend account?

No. It uses the public anonymous review flow.

Does it download full review text?

V1 returns the title and teaser exposed on product review cards, plus the direct full-review URL.

Can I scrape several products in one run?

Yes. Add multiple entries to startUrls; maxItems applies globally.

Can I collect only negative reviews?

Yes. Set maxRating to 1 or 2.

Can I collect reviews from a date range?

Yes. Use inclusive dateFrom and dateTo values.

Why did I receive fewer rows than maxItems?

Filters may exclude cards, the source may have fewer reviews, or iRecommend may have stopped serving a later page. Check the run log for the exact page outcome.

Why is imageUrls shorter than photosCount?

Cards expose only preview images. photosCount reports the review's total visible photo count.

Why does the first page briefly return 521?

That can be iRecommend's anonymous JavaScript verification response. The Actor waits for the automatic reload and real review cards.

How do I avoid duplicate records between scheduled runs?

Use reviewId as your downstream upsert or deduplication key.

Related scrapers

Explore other public review and product-data tools from Automation Lab:

• Trustpilot Scraper
• Google Maps Reviews Scraper
• Amazon Reviews Scraper

Choose the source-specific Actor that matches the website your workflow monitors.

Support

If a public iRecommend product page stops producing expected review rows:

1. Keep the smallest reproducible input.
2. Note the source URL and run ID.
3. Compare the visible cards with the dataset.
4. Open an Actor issue with those details.

Clear reproduction details help distinguish a changed selector, source outage, or route-specific verification problem.