TrustRadius Software Scraper

Scrapes B2B software rankings and reviews from TrustRadius — a G2/Capterra competitor — by category.

Give it a category (crm, project-management, hr-management, ...). Get back the top-rated products with pricing, awards, pros/cons, a sample of full-text reviews, and reviewer demographics.

---

What you get

One row per product, plus a CATEGORY_SUMMARY row per category.

From the category page (always)

name · url · categoryRank · image · brand · plus the category's own aggregateRating (rating value and review count as TrustRadius shows it on the listing card).

From the product page (includeProductDetails, on by default)

startingPrice / priceCurrency · awards (e.g. "Top Rated 2026") · ratingValue · reviewCountTotal (the real total review count) · reviewsEmbedded + the actual review text/author/date in productDetailJsonLd.review · positiveNotes / negativeNotes (pros/cons) · topIndustries / companySizes / topCountries / departments (who actually uses this product, from TrustRadius' own reviewer demographics).

---

Read this before you size a crawl: the category list is capped at 25

TrustRadius' category page always renders its top 25 products by rating — and that's it. Every pagination and sort parameter tried (page, offset, start, sort=rating/reviews/az) returned the exact same 25 products, verified by comparing the actual product URLs, not just counts. The real listing mechanism lives behind TrustRadius' own API, which robots.txt closes off.

This isn't hidden in the data: CATEGORY_SUMMARY.categoryTotalProducts reports the category's real size (CRM: 1,554 products) right next to productsReturned (never more than 25) and cappedBelowTotal (true whenever the category has more than this actor can reach). For most purposes this is still the useful slice — nobody wants product #950 of 1,554 in a software category — but size your expectations accordingly.

---

Input

{
  "categories": ["crm"],
  "maxItems": 25,
  "includeProductDetails": true
}

A category slug that doesn't exist answers a clean 404 — one ERROR row, never wrong data — so slugs aren't pre-validated against a fixed list (there are hundreds, and TrustRadius' own category-page URL is the source of truth).

---

Known limits

Review counts use different field names on different pages, and this actor already shipped that bug once. The category page's own rating block uses reviewCount; the product detail page's uses ratingCount — the exact same number, two different Schema.org property names, verified on the same product returning 283 both ways. An early version of this actor only read one of them and silently returned null for every single row. reviewCountTotal now checks both.

The embedded reviews are a small sample, not the full set. A product with reviewCountTotal: 283 typically ships around 5 reviews embedded in its own page — reviewsEmbedded tells you how many actually came through.

---

Anti-bot posture

A TLS-fingerprint gate: of 5 profiles tried, 3 came back clean (chrome124, safari17_0, firefox133) and 2 (chrome99_android, edge101) answered HTTP 403. This actor's profile pool avoids both blocked ones. Residential proxy is the cloud default regardless.

Policy

robots.txt: Allow: /, then Disallow: /api/, /share/, /search/, /profile/, /vendoradmin/, /cdn-cgi/. Category pages (/crm, ...) and product pages (/products/{slug}) are not disallowed — this actor never touches /api/, /search/ or any other closed path. No ClaudeBot/ anthropic-ai group.