Trustpilot Reviews Scraper — Full History, Past the 200 Cap

Export Trustpilot reviews for any company — far beyond the 200 reviews Trustpilot shows per filter — with full text, star rating, company replies, reviewer country and verification labels, plus a company profile row with TrustScore, star histogram, language breakdown, categories and public contact details.

Every run tells you exactly how much of the company it got: each company row carries reviewsCollected, coveragePct and stopReason, and the COVERAGE record summarises every input.

What you get

📝 Review rows	title, full text, 1–5 stars, language, likes, published / experienced / updated dates, source (organic, invited…), verification level
💬 Company replies	reply text, reply date, hours-to-reply
👤 Reviewer	name, country, profile link, avatar, total reviews written, reviews on the same company
🏢 Company profile	TrustScore, stars, total reviews, reviews in last 12 months, 1–5★ histogram, per-language counts, categories, claimed/verified flags, consumer alerts, email/phone/address, reply rate & speed, Trustpilot AI summary
📊 Coverage	per company: reviews on Trustpilot matching your filters, reviews collected, % coverage, why it stopped

How deep does it go?

Trustpilot's website stops at page 10 (200 reviews) for any filter combination. This actor asks for the same company in thousands of different, narrower ways — per language, per star rating, with/without replies, and by review keywords learned from the reviews themselves — and merges the results without duplicates.

Two ways to go past 200:

• Newest first (default). You get the most recent reviews, newest first, in date order. For example, the newest 1,000 for amazon.com run from today back to late August. Splitting by language and star rating is exact, so every review newer than the company row's completeSince is guaranteed to be included. Further back, the actor digs into the busiest groups (e.g. English 1★) with keyword and relevance queries, best effort. Measured on amazon.com, that reached 746 of the 750 reviews from the last 30 days. With max reviews 0, it collects as deep as deep history does, then delivers newest first.
• Deep history (deepHistory: true). This collects as much of a company's history as possible. Rows arrive as found, not in date order.

Deep history, measured on the Apify platform on 30 Sep 2026 (default datacenter proxy, no cap):

Company	Reviews on Trustpilot	Collected	Time	Oldest review
nordvpn.com	50,744	36,521 (73 %)	13 min	2015
bitdefender.com	12,540	11,516 (92 %; 97 % of the reviews Trustpilot lists publicly)	5 min	2007
nordvpn.com, 1★ in German	497	479 (96 %)	1 min	2019

With a cap, 3,000 nordvpn reviews take about 45 seconds. Other Trustpilot scrapers tested the same day stopped at 200–3,000 reviews for nordvpn, and two returned nothing.

Input

{
  "startUrls": ["nordvpn.com", "https://www.trustpilot.com/review/www.bitdefender.com?stars=1"],
  "maxReviewsPerCompany": 1000,
  "sort": "recency",
  "languages": ["all"]
}

• Companies — domains or Trustpilot review URLs. Filters in a pasted URL are applied to that company.
• Max reviews per company — default 1,000; 0 = as many as reachable.
• Order (sort):
  • recency (default): the newest reviews in exact date order, newest first, also beyond 200.
  • relevance: Trustpilot's "most relevant" order. That order only exists for the first 200 per filter; beyond that rows are added for coverage, unordered.
• Deep history (deepHistory) — maximum coverage of a company's whole history. Rows are not in date order. Best for full exports with max reviews 0.
• Filters — stars, languages (default all), published within, verified only, with company reply only, keyword ("quoted phrase" for exact match). Filters are applied to every request, so filtered runs still go past 200.
• Only reviews newer than — monitoring mode for scheduled runs: newest-first, stops at your date.
• Company profile row — on by default.

Output

One dataset; type is review or company. Every field is always present (null when Trustpilot has no value).

completeSince (company row, newest-first runs): every review matching your filters and published at or after this time is in the dataset. It equals the oldest returned review when no language × star group overflowed Trustpilot's 200 cap. It is more recent when a very busy group had to be dug into; the rows before that point are still the newest the actor found, in date order, collected best effort. It is null when nothing can be guaranteed, for example after failed requests, with no cap, or with relevance order. It is null for relevance, deep-history and monitoring runs. Two views: Reviews and Companies. The key-value store record COVERAGE lists every input with status (ok, not-found, failed), coverage and stop reason.

stopReason values:

• complete — everything matching your filters was collected
• cap — reached your max reviews per company
• diminishing / exhausted / budget — the remaining reviews are not reachable at reasonable cost
• depth-limit — (monitoring) the window holds more reviews than Trustpilot exposes
• charge-limit — your spending limit was reached
• not-found — no Trustpilot page for that domain (not charged)
• invalid-filter — Trustpilot rejected one of your filters
• blocked / error — Trustpilot could not be reached reliably; partial results are kept
• timeout — the run was about to hit its time limit; everything collected so far is delivered

Pricing

Pay per result: $0.75 per 1,000 reviews and $0.003 per company profile. No start fee. Duplicates, companies that don't exist on Trustpilot and failed requests are never charged. Set a spending limit on the run and the actor stops cleanly when it is reached.

FAQ

Do I need a Trustpilot account or API key? No. Only public review pages are used.

Why not 100 % coverage? Trustpilot hides reviews beyond page 10 of every view. Coverage depends on how varied a company's reviews are; the company row tells you exactly what you got.

Which proxy? The default Apify datacenter proxy works; residential is used automatically only if datacenter IPs get blocked.

Is scraping Trustpilot legal? The actor reads publicly visible review pages. Reviewer names are personal data under GDPR — make sure you have a lawful basis for how you store and use them.