Capterra Reviews Scraper: Pros, Cons, Company Size

Give it Capterra product URLs and it returns one JSON row per review: overall and sub-ratings, pros, cons, review text, the reviewer's role, industry and company size, how long they used the product and the incentivized flag. $1.79 per 1,000 reviews on the free plan, and you pay only for the rows you get.

Why this actor

• Capterra's own sort and filters (stars, company size, role, usage) run server-side, so the rows are exactly the slice you asked for.
• Up to 50,000 reviews per product, past the 2,500 you reach by clicking through the site.
• Native review IDs for deduplication, and onlyNewReviews for weekly monitoring.
• Input names of other Capterra review scrapers (productUrl, maxResults, sort, starRating) work as they are. Free error rows with a machine-readable code (not_found, blocked, unavailable...), so an agent knows whether to retry.

Built by Arcadium Data: public data, dated and agent-ready, never personal data.

Use with AI agents

Returns Capterra reviews (pros, cons, rating, reviewer role and company size) for software product URLs. Agents can call it through the Apify MCP server (https://mcp.apify.com, open source as apify/actors-mcp-server) as arcadium/capterra-reviews: the input schema is read from the actor, the result is a dataset of JSON rows.

Input example (minimal):

{
  "startUrls": [
    {
      "url": "https://www.capterra.com/p/135003/Slack/reviews/"
    }
  ],
  "maxReviewsPerProduct": 20
}

What comes back: one review row per review (charged), free coverage rows (stated total vs delivered) and a free error row for an input that could not be read. Filter on the rowType field. An input that fails is reported, not hidden (includeErrorRows, on by default): the error row carries rowType: "error", a code (invalid_input, not_found, blocked, unavailable, filter_unsupported or no_match), a message and the input it is about. Retry on blocked and unavailable later; do not retry the other codes with the same input.

Cost of the call above: 20 reviews at $1.79 per 1,000 plus the $0.001 start fee is about $0.04 on the free plan; paid plans pay less. Free rows are never charged.

What you get per review

Output example:

{
  "rowType": "review",
  "platform": "capterra",
  "productName": "Slack",
  "productUrl": "https://www.capterra.com/p/135003/Slack/",
  "entityId": "135003",
  "reviewId": "Capterra___7224147",
  "reviewUrl": "https://www.capterra.com/p/135003/Slack/#Capterra___7224147",
  "title": "Very strong communication tool.",
  "text": "Generally very good, quick and easy. Good options to integrate, easy to file share. I typically run multiple Slack instances.",
  "pros": "Ability to communicate in real time with colleagues when we're working remotely. Also for file sharing.",
  "cons": "The threads feature is not something that I find particularly intuitive...",
  "ratingOverall": 5,
  "ratings": { "easeOfUse": 5, "customerService": 4, "features": 5, "valueForMoney": 5, "likelihoodToRecommend": 9 },
  "date": "2026-10-02",
  "reviewerRole": "CPO",
  "reviewerIndustry": "Computer Software",
  "reviewerCompanySize": "1-10 employees",
  "reviewerTimeUsed": "1-2 years",
  "switchedFrom": "Google Workspace",
  "reasonForSwitching": "Google Workspace chat works well. However, several of our technical staff wanted to use Slack.",
  "incentivized": true,
  "incentive": "NominalGift",
  "reviewerVerified": true,
  "product": {
    "platform": "capterra",
    "entityId": "135003",
    "name": "Slack",
    "url": "https://www.capterra.com/p/135003/Slack/",
    "rating": 4.7,
    "reviewCount": 24229,
    "category": "Team Communication",
    "vendor": "Slack",
    "website": null
  },
  "scrapedAt": "2026-10-20T17:49:58.237Z"
}

A few fields worth knowing about:

reviewId is Capterra's own ID, so it stays the same between runs. Use it to dedupe when you scrape the same product every week.

incentivized tells you whether Capterra marked the review as written for a gift card or similar reward. Plenty are. If you're doing sentiment analysis you probably want to look at both groups separately.

switchedFrom, alternativesConsidered and reasonForSwitching are filled only when the reviewer answered those questions, which is maybe one review in four. They're the most useful fields for competitive research, so the gaps are a pity, but that's the data.

product repeats on every review, so each row stands on its own when you load it into a spreadsheet or a database. The same product, plus pricing plans, alternatives, features and integrations, is saved once per product in the PRODUCT record of the run's key-value store.

Sub-ratings are left out when the reviewer skipped them (you won't see a fake 0). likelihoodToRecommend is on a 0 to 10 scale; the rest are 1 to 5.

Reviewer names and profile photos are never collected. You get the role and company details, not the person.

Input

Give it startUrls. Any Capterra product link works, with or without /reviews/ at the end, and the regional Capterra sites (capterra.co.uk, capterra.com.au and so on) are accepted too:

{
  "startUrls": [
    { "url": "https://www.capterra.com/p/135003/Slack/" },
    { "url": "https://www.capterra.com/p/186596/Notion/reviews/" }
  ],
  "maxReviewsPerProduct": 500,
  "sort": "MOST_RECENT"
}

If you call the actor through the API, note the { "url": ... } objects. A plain list of strings gets rejected by Apify's input validation.

Everything else is optional:

Field	What it does
maxReviewsPerProduct	How many reviews to collect per product. Default 100, max 50,000.
sort	Most recent, most helpful, highest rated or lowest rated. Same options as the site.
minRating, maxRating	Overall star range, for example 1 to 2 if you only want complaints.
dateFrom	Only reviews from this date on (YYYY-MM-DD). With "Most recent" the run stops as soon as it reaches older reviews, so you don't pay for anything you'll throw away.
companySizes	Self-employed, 1-10, 11-50 and so on up to 10,001+ employees.
reviewerRoles	User, administrator or other.
timeUsed	Free trial, under 6 months, 6-12 months, 1-2 years, 2+ years.
usageFrequency	Daily, weekly, monthly, other.
includeProductDetails	On by default. Adds the product's exact name, rating, category and vendor to every review and saves the full profile (pricing, alternatives, features, integrations) in PRODUCT. Turn it off if you only care about the reviews; the name is then taken from the URL.
onlyNewReviews, stateKey	Incremental mode for scheduled runs. Only reviews that earlier runs with the same stateKey did not deliver are returned, and only those are charged. Uses the "Most recent" sort. The state lives in a key-value store named capterra-reviews-seen in your account; pick a new stateKey to start from scratch.
includeErrorRows	On by default. An input that can't be scraped (bad URL, product not found) produces one row with "rowType": "error" and a code, and the run still succeeds. Error rows are never charged.
includeCoverageRows	On by default. One free row per product with "rowType": "coverage": the review total the site states, how many were delivered, why the run stopped and whether anything is missing (shortfall). Never charged.
markEdited	On by default, with onlyNewReviews. Reviews seen before whose text changed in the last 7 days are delivered again with extras.changeType set to edited (charged like a new review). New ones carry new.

Filters are applied by Capterra before anything is downloaded. Ask for 1-star reviews from companies with 1-10 employees and you'll get exactly those, without paging through thousands of others first.

How many reviews can I get?

All of them, up to 50,000 per product. There's no 2,500 cap like the one you hit when clicking through pages on the site. In our tests Slack reviews came through all the way back to 2017. Most products have far fewer reviews, so this mostly matters for the big names.

A run of 1,000 reviews usually takes under a minute.

Coming from zen-studio's Capterra scraper?

These input names work as they are: productUrl (one link), maxResults (0 means no limit), sort (including HIGHEST_COMPLETENESS_SCORE, which becomes "most helpful") and starRating (a list such as ["1","2"], applied as a min to max range). feedback is ignored. The first time one of these names is used, the run log says which of our fields it was mapped to.

Pricing

You pay per review delivered, at the price of your Apify plan. Coverage rows, error rows and repeats in only-new mode are free.

Apify plan	Per 1,000 reviews
Free	$1.79
Starter (Bronze)	$1.34
Scale (Silver)	$1.25
Business (Gold)	$1.16
Platinum	$1.16
Diamond	$1.16

Plus a one-off $0.001 per GB of memory when a run starts.

Set a maximum cost per run in the Apify run options and the actor stops before it goes over.

Related actors and data

• Software Reviews Scraper: Capterra, Google Play, App Store
• Capterra Products Scraper: Pricing, Features, Rank
• TrustRadius Reviews Scraper: Pros, Cons, Reviewer Role
• Review Monitor: Negative Review Alerts and Rating Drift
• Arcadium Data MCP Server: AI agents can ask the same questions as tools (search_reviews, review_snapshot).

Limitations

Capterra doesn't publish a stable public API, so the site can change without warning. We run checks against the live site every day and usually notice breakage before you do. If Capterra changes something big, the actor keeps working in a reduced mode (default order, up to 2,500 reviews per product) while we fix it. In that mode the reviewer role and usage frequency filters can't be applied, so a product that uses them fails with a clear message instead of returning the wrong reviews.

helpfulVotes and language are part of our shared review format but Capterra doesn't expose them, so they're always empty here. reviewUrl is the product URL plus the review ID as an anchor; Capterra has no page of its own for a single review.

The vendor website isn't published in the data we read, so product.website is empty. Reviews from the regional Capterra sites are read from capterra.com, which lists the same products and reviews.

In the reduced mode described above, onlyNewReviews can't rely on newest-first order and may miss new reviews that sit deeper than the first pages. It goes back to normal as soon as the standard mode works again.

Vendor replies (vendorResponse) are rare on Capterra. We haven't seen one in our test data yet, so treat that field as best effort.

Is this allowed?

The actor only reads reviews that anyone can see on Capterra without logging in, and it doesn't collect personal data about reviewers. How you use the data is up to you. If you plan to republish review text, check Capterra's terms first.

Questions or problems

Open an issue on the Issues tab with the product URL and your input. If something's broken we'd rather hear about it than have you quietly give up on the actor.