Trustpilot Review Scraper: Collect Customer Reviews from Any Company Page

Pull customer reviews from any Trustpilot company profile. Give it a domain like www.amazon.com and it walks through the review pages collecting star ratings, full review text, reviewer details, and any company replies. No extra setup beyond the domain, and it handles pagination on its own.

Use cases

• Reputation monitoring: track what customers say about your brand over time without checking the site manually
• Competitor research: collect reviews for rival businesses and compare sentiment, common complaints, and rating trends
• Sentiment analysis: feed structured review datasets into NLP pipelines for opinion mining and topic modeling
• SEO research: identify keywords and themes customers use when talking about a brand to inform content strategy
• Market research: aggregate review data from multiple companies in an industry to spot patterns and benchmark performance

What data does this actor extract?

Each review record includes:

{
    "reviewId": "abc123def456",
    "companyDomain": "www.amazon.com",
    "companyName": "Amazon",
    "reviewTitle": "Fast delivery but packaging was damaged",
    "reviewText": "Ordered a kitchen appliance and it arrived two days early. The outer packaging was crushed and the product had a small scratch.",
    "rating": 3,
    "reviewDate": "2024-11-15T09:23:00.000Z",
    "reviewerName": "Sarah M.",
    "reviewerCountry": "GB",
    "verified": true,
    "companyReply": "Thank you for your feedback. Please contact our support team and we will make this right.",
    "scrapedAt": "2025-03-01T14:22:10.123456+00:00"
}

Field	Type	Description
reviewId	string	Unique Trustpilot identifier for the review
companyDomain	string	Company domain used as input
companyName	string	Company display name as shown on Trustpilot
reviewTitle	string	Review headline written by the customer
reviewText	string	Full body text of the review
rating	integer	Star rating from 1 (worst) to 5 (best)
reviewDate	string	ISO 8601 date when the review was published
reviewerName	string	Display name of the reviewer
reviewerCountry	string	ISO 3166-1 alpha-2 country code of the reviewer
verified	boolean	Whether Trustpilot verified the reviewer made a real purchase
companyReply	string	Company response to the review. Empty string if no reply was posted
scrapedAt	string	ISO 8601 timestamp of when this record was collected

Input

Parameter	Type	Default	Description
companyDomain	string		Company domain as registered on Trustpilot (e.g. www.amazon.com) or full Trustpilot URL. Required.
maxReviews	integer	100	Maximum reviews to collect per run. Capped at 1000.
startPage	integer	1	Page to start from. Set higher to skip earlier pages.
endPage	integer	0	Page to stop at (inclusive). 0 = no page limit, use maxReviews instead.
stars	string	All ratings	Filter by star rating: 5, 4, 3, 2, or 1. Leave on default for all ratings.
datePosted	string	All reviews	Filter by date period: All reviews, Last 30 days, Last 3 months, Last 6 months, Last 12 months.
language	string	All languages	Filter by language: English, Deutsch, Français, Español, Italiano, Nederlands, and more.
requestTimeoutSecs	integer	30	Per-request timeout in seconds.
proxyConfiguration	object	Datacenter (Anywhere)	Proxy type and location for requests. Supports Datacenter, Residential, Special, and custom proxies. Optional.

Example input

{
    "companyDomain": "www.amazon.com",
    "maxReviews": 200,
    "language": "en",
    "stars": 1,
    "proxyConfiguration": { "useApifyProxy": true }
}

You can also pass a full Trustpilot URL instead of a bare domain:

{
    "companyDomain": "https://www.trustpilot.com/review/www.booking.com",
    "maxReviews": 50
}

To scrape a specific page range (useful for splitting large runs across multiple actor invocations):

{
    "companyDomain": "www.amazon.com",
    "startPage": 6,
    "endPage": 10,
    "maxReviews": 1000
}

How it works

1. The actor normalizes the input domain. Trustpilot company domains often include www. (e.g. www.amazon.com), so the prefix is preserved as-is.
2. Page 1 — HTML fetch: the actor sends a browser-realistic request (Chrome 145, full sec-ch-ua/sec-fetch-* headers, Google as referrer) and parses the __NEXT_DATA__ JSON script tag embedded in the HTML. This gives the first 20 reviews and, critically, the Next.js buildId needed for pagination.
3. Pages 2+ — Next.js JSON API: instead of fetching full HTML pages, the actor calls Trustpilot's internal data API directly (/_next/data/{buildId}/review/{domain}.json?page=N&businessUnit={domain}). This endpoint returns pure JSON with the same review structure, no HTML parsing needed. The request uses x-nextjs-data: 1, sec-fetch-mode: cors, and sec-fetch-dest: empty to match what the browser sends.
4. Reviews are parsed from each response and pushed to the dataset. The actor repeats until it reaches maxReviews or the API returns an empty reviews list.
5. If proxy configuration is set, a fresh proxy URL is obtained for each page request.

Integrations

Trustpilot Review Scraper works with Apify integrations including Google Sheets, Slack, Make, and Zapier. Use webhooks to trigger downstream actions when a run finishes and new data is ready.

FAQ

Can this actor scrape reviews for any company on Trustpilot? Yes. As long as the company has a public Trustpilot profile, pass the company domain or the full Trustpilot URL as input.

What domain format should I use: amazon.com or www.amazon.com? Use the format that matches the company's Trustpilot URL. Many large companies (including Amazon) are registered as www.amazon.com on Trustpilot. If you get zero results with amazon.com, try www.amazon.com.

How many reviews can I collect per run? The actor caps at 1000 reviews per run to keep costs predictable. For larger datasets, run the actor multiple times with different star rating or language filters and merge the results.

Does it need a Trustpilot account or API key? No. It scrapes public review pages without any authentication.

Will it collect reviews in languages other than English? Yes. Leave language empty to get reviews in all languages, or set it to a language code like de or fr to filter.

What should I do if the actor gets blocked? Switch to Residential proxies in the proxy configuration. Datacenter IPs are faster but more likely to trigger Trustpilot rate limits.

How many reviews are on each page? Trustpilot shows 20 reviews per page, so collecting 100 reviews requires roughly 5 page requests.

Legal notice

Trustpilot review pages are publicly accessible, but check Trustpilot's terms of service and your local regulations before using scraped data commercially.