Capterra Reviews Scraper

A software product's Capterra reviews, exported as a spreadsheet. Every row carries the reviewer's headline, their overall comment, and — kept as separate columns rather than run together — what they liked and what they disliked. Beside the prose sit the five per-criterion scores behind the star rating, the would-you-recommend score, and the three things that make a review set segmentable: the reviewer's job title, their industry, and how long they actually used the product.

Pros and cons being apart is the part that matters in practice. A search for complaints across one block of concatenated review prose also matches the praise in the same sentence; a search across cons does not.

Accepted input

productUrls is required and takes one or more Capterra product pages, one per line, as full links such as https://www.capterra.com/p/147657/monday-com/ or as the pair inside one, 147657/monday-com. The number on its own is not enough — a Capterra product is addressed by a number and a name, the name is case sensitive (monday-com, but GitLab and Structure), and neither half can be worked out from the other. Every row Capterra Category Scraper produces carries the whole address in url, so a category export feeds straight in.

maxItems bounds each product separately and defaults to 100. Reviews arrive 25 to a page, so 100 is four pages per product. The cap is per product rather than per run because a run given twenty products is asking for a comparable slice of each. Set 0 to walk a product to the end.

startPage sets the page each product begins at and defaults to 1. It resumes a walk that stopped and splits a long product across several runs. Its ceiling is 100, for the reason in the next section.

{
  "productUrls": ["https://www.capterra.com/p/147657/monday-com/"],
  "maxItems": 25,
  "startPage": 1
}

Response fields

One row per review. This is the first row of that exact run, with the three prose fields shortened for length:

{
  "product_id": "147657",
  "slug": "monday-com",
  "page": 1,
  "sort": "most helpful reviews",
  "total_reviews": 6095,
  "page_limit": 100,
  "product_name": "monday.com AI Work Platform",
  "product_rating": 4.6,
  "product_rating_best": 5,
  "product_review_count": 6095,
  "title": "A Powerful AI Assistant That Streamlines Project Management and Boosts Productivity",
  "body": "My experience with the monday.com AI Work Platform has been overwhelmingly positive. It has significantly enha ...",
  "pros": "he monday.com AI assistant has completely transformed my workflow management. With simple text prompts, I can  ...",
  "cons": "The AI features in monday.com have felt underwhelming and gimmicky for our team's daily operations. The automa ...",
  "date": "August 8, 2026",
  "source": "Incentivized review",
  "review_rating": 5,
  "review_rating_best": 5,
  "criteria_ratings": {
    "Overall Rating": 5,
    "Ease of Use": 5,
    "Customer Service": 5,
    "Features": 5,
    "Value for Money": 5
  },
  "likelihood_to_recommend": 10,
  "author": "Marquita V.",
  "role": "SALES MANAGER",
  "industry": "Biotechnology",
  "used_for": "1-2 years"
}

title and body are genuinely two fields on this site rather than one field formatted twice. They are published separately because they are separate.

criteria_ratings is Capterra's own map, keyed by its own labels, rather than five fixed columns. A reviewer who skipped Customer Service has no key for it — a column would publish that as a score of nothing, which is a different claim.

likelihood_to_recommend is out of 10. Every other score on the row is out of 5, and review_rating_best sits beside the star rating so the two axes stay apart.

sort says what arrived, not what was asked for. Capterra has no sort parameter at all — its sort control does not put anything in the address — so reviews come in its most-helpful order and the order is reported rather than chosen. There is deliberately no input pretending otherwise.

role is printed as the reviewer typed it, capitals included.

Behaviour on partial results

A product address Capterra serves nothing for is logged and skipped, and the run continues with the next one. A product that exists with no reviews returns a successful run with no rows for it.

At most 2,500 reviews of any product are reachable, whatever total_reviews says. Capterra paginates every product to page 100 and no further — the product above states 6,095 reviews and stops at page 100, so 2,500 are addressable and the other 3,595 cannot be opened from the site at all, by this Actor or by anyone in a browser. The walk stops there and the run log says so when a product is bigger than its own pagination. startPage is capped at 100 for the same reason: a higher number has no page behind it.

Pages do not overlap. Measured on 2026-08-23, pages 1 and 2 of the product above came back as stated ranges 1–25 and 26–50 and shared no review headline.

Frequently asked questions

Can I sort by newest, or by lowest rating? No, and no Actor can — Capterra's own sort control posts a request back to the page rather than putting a sort in the address, so there is no sorted address to fetch. Every run gets most-helpful order, and sort on every row states it so a dataset is never assumed to be recency-ordered.

Why does total_reviews say 6,095 when the run stopped at 2,500? Because those are two different numbers. 6,095 is Capterra's count of the reviews a product has; 2,500 is how many its pagination will hand over. page_limit on every row is the honest one.

Why did a page fail and then work on its own? Capterra sometimes refuses a request and states how long to wait. Two of eight live calls on 2026-08-23 came back that way. The wait is honoured and the page asked for again; the run reports it and continues rather than failing.

Where do I get product addresses in bulk? From Capterra Category Scraper, which walks a category in ranked order and publishes each product's address as url. Guessing one does not work — the name half is Capterra's spelling rather than the product's.

Is a Capterra account, cookie or API key required? No. No account, session cookie or key of any kind is supplied to the Actor or needed by it.

Related Actors

Capterra Category Scraper is where product addresses come from: a category in ranked order, with each product's rating, review count, summary and Shortlist flag.

Capterra Profile Scraper reads the vendor's own description, the supported platforms and the product Q&A — none of which appear on a reviews page.