Platform guides

Capterra reviews: how the platform works for software vendors

1 October 2026 · JustReview

Capterra reviews are verified user reviews of business software, published by Gartner on three sites at once: Capterra, GetApp and Software Advice. A vendor lists a product for free and pays Gartner for clicks or leads, while reviews arrive either on their own or because Gartner invited the reviewer, sometimes with a small incentive that is disclosed on the card. For a vendor that means one review count that follows the product across three storefronts and a rating that is hard to inflate, but also an archive that goes stale the moment nobody keeps asking.

That is the theory. To see the practice, we read every one of the 1,133 public reviews of one shipping-software product on Software Advice on 1 October 2026, checked the same product on GetApp, and compared both with the copy we imported from its Capterra page for a JustReview account. The product has an ordinary profile for its category: a 4.8 rating, a long history and a customer base of small online sellers. What it shows is how the platform behaves when you look at all of a vendor’s reviews instead of the first page.

How Capterra reviews work, rule by rule

Gartner publishes its review rules in the guidelines linked from every page footer, and the Software Advice reviews page sums up the verification promise in one paragraph: human moderators check that reviewers are real people and that reviews are authentic, and they screen the text for plagiarism and generated content. Each rule below sits next to what 1,133 real reviews show.

Two rows deserve more than a table cell, because they change how a vendor should think about the platform: the incentive label and the disappearing reviews.

One review, three storefronts

The first thing a vendor should understand is that “my Capterra reviews” is a slightly misleading phrase. Gartner owns all three directories, and a review collected on any of them is shown on all of them. On the product we measured, GetApp and Software Advice both reported 1,133 reviews with the same five-to-one-star split down to the single two-star review. Our import carries the trail: reviews read from the Capterra page have identifiers that start with the name of the site where they were written, and a one-star review posted on Software Advice in April 2025 sits on the Capterra product page as SoftwareAdvice___6780734. GetApp marks the same review with Software Advice as its source.

The practical consequence is pleasant. When you ask a customer for a review, you can send them to whichever of the three they find easiest, and the review lands on all three. It also means there is no point running separate review campaigns per directory, or reporting the three totals as if they were independent proof. They are one number displayed three times.

How many Capterra reviews are incentivized

Software Advice prints the origin of every review in a “Review source” note on the card, and there are only two versions of it. One reads “This review was submitted organically. No incentive was offered”. The other reads “This reviewer was invited by us to submit an honest review and offered a nominal incentive as a thank you.” Counted across all 1,133 reviews of our product, 923 carry the first and 210 the second, so 18.5% of the profile exists because Gartner went looking for reviewers.

The pattern over time says more than the total. Invited reviews come in waves: Gartner’s outreach brought 69 reviews in 2018 and 62 in 2019, fell to 17 and then 7, and returned in 2022 with 23. Organic reviews peaked a year after the first wave, at 277 in 2019, and then collapsed, from 129 in 2020 to 4 in 2021. Since then invitations have carried almost half of the profile: 46 of the 95 reviews written from 2021 to 2025 were invited, against 15.8% of the 1,038 written before. Without Gartner’s outreach, this product would have collected 49 reviews in five years.

The other thing the count settles is whether invited reviewers are softer graders. They are not, if anything they are cooler. 73.8% of the incentivized reviews gave five stars against 86.2% of the organic ones, the average rating is 4.67 against 4.83, and three stars are almost five times as common among the invited, 6.7% against 1.4%. What the invited group lacks is the angry end of the scale: all six one- and two-star reviews on the profile were organic. On this product, an incentive that does not depend on the rating bought a calmer verdict, not a better one. If you display these reviews on your own site, keep the disclosure with them; the US rules on that are summarised in our guide to FTC rules for displaying customer reviews.

Reviews that quietly disappear

Our copy of this product’s Capterra reviews holds 1,148 entries, almost all of them imported in January 2024. The live network shows 1,133. The difference is not spread evenly. The one-, two- and three-star counts match exactly, 5, 1 and 27 in both places, while our copy has 12 more five-star and 3 more four-star reviews than Gartner now publishes. The extra reviews are not one batch: they were written between 2015 and 2023, in a dozen different months, and all of them were on the public page when we imported it.

We cannot see why a review was withdrawn, and there are innocent explanations: a reviewer can ask for removal, and a moderator can pull a review that fails a later check. What the numbers do show is that removals on this profile only ever cost the vendor praise, never criticism. That is the opposite of what a vendor fears when they think about moderation, and it is one more reason the count on your own site will drift from the count on the directory. The other usual reasons are in why review counts do not match.

The archive problem

A profile with more than a thousand reviews looks permanent. It is not, because the date on the newest review is what a buyer reads first. On the product we measured, the newest review in the whole network was written on 19 November 2025, and it is a one-star complaint about a refund. Our import has the same review as its newest, so the silence is on the platform, not in our connector. More than ten months without a single new review is a long silence for a product that once collected hundreds a year.

Our own book of Capterra reviews tells the same story in a different way, and it needs a correction. In September we reported that our Capterra reviews peaked in 2019 and almost stopped after 2022. Today’s check shows that 5,449 of the 6,697 Capterra reviews we hold belong to one account whose import ended in August 2022, so part of that curve measured when we stopped reading a profile rather than when Capterra stopped collecting. The product above is the cleaner test, and it still shows a real decline: 339 reviews written in 2019, 22 in 2023, 14 in 2024 and 4 in 2025. A large archive and a product that buyers are reviewing today are two different things, and a directory shows the first more prominently than the second.

How to get Capterra reviews without breaking the rules

Gartner runs its own outreach, so part of your review count arrives without your involvement. The part you control works best when you treat the three directories as one destination:

• Ask right after a visible success, such as a first month closed or an integration that went live, not inside a renewal email. That is when a customer has something specific to say.
• Send one link to whichever directory the customer already uses. As shown above, the review appears on all three.
• Never make a reward depend on the rating or on the review being published. Gartner discloses its own incentives on every card, and an undisclosed one from you is exactly what the guidelines and the FTC forbid. Our piece on why buying reviews is a trap covers what detection looks like.
• Answer critical reviews in public. Our product has 6 vendor responses across 1,133 reviews, the newest from January 2022, and only one of its six one- and two-star reviews has an answer that a later buyer can read.

What a Capterra review is worth on your own website

The badge Capterra offers vendors is an image that links back to the directory, so it sends a buyer away at the moment they were deciding. The review itself is better material than a badge. It names the reviewer’s company size and how long they have used the product: on our product, 597 of the 797 reviewers who stated a company size, 74.9%, work alone or in a team of 2 to 10, and 218 of the 920 who described their usage had used the software daily for more than two years. That is the context a buyer from a similar company is looking for, and it belongs next to your price, as we argued in social proof on a SaaS pricing page.

Our Capterra reviews widget reads your public product page from the address you paste in and renders the reviews as text on your own domain, with the Capterra mark on every card so a sceptical buyer knows where to check them. Because the network shares one pool, the Capterra page also carries the reviews your customers wrote on GetApp and Software Advice. The setup, including which URL to paste, is in how to add Capterra reviews to your website.

Capterra is part of the Ultimate plan at EUR 39 a month, which also covers every other source we support and up to 20,000 stored reviews, with a 7-day free trial; the full ladder is on the pricing page. Open an account, paste your Capterra product page, and check the imported count against the number Capterra shows before you choose a layout. If the two differ, you now know where to look first.

How this was measured

On 1 October 2026 we loaded all 46 pages of public reviews for one shipping-software product on Software Advice, 25 reviews per page, and recorded the review date, star rating, review-source note, reviewer company size, usage line and whether a vendor response or update was shown. We kept no reviewer names or review text beyond what was needed to match records. The same day we read the product’s GetApp listing for the total and the star split, and compared everything with the 1,148 reviews our database holds from its Capterra page. One product is a case, not a census, so treat the shares as an example of how the platform works rather than a market average.