ReviewMeta: what happened to it

It never claimed to know which individual review was fake. It ran statistical tests over the whole review set: how many reviewers had only ever written one review, whether ratings arrived in unnatural bursts, how much of the wording was duplicated across reviews, how the rating distribution compared to normal products. Each test contributed to a pass, warn or fail, and the failures were stripped out to produce the adjusted rating.

That framing was its strength. It gave shoppers and sellers a defensible read on a listing without accusing anyone of anything, and it was free, which is why it spread through browser extensions and Reddit threads rather than through marketing.

On 11 January 2021 the founder published a post looking for a successor. He explained that he had been running the project largely alone for years, that the work had stopped being sustainable for him personally, and that he wanted someone else to carry it forward. The post is still readable in the Internet Archive.

No successor took it over in any visible way. The service limped on and then stopped. Today reviewmeta.com does not serve the application at all: the domain answers with a bare server error rather than a page. We last checked on 3 September 2026.

We are not going to put a precise shutdown date on this. The public record shows the 2021 handover attempt and it shows the site being unreachable now. The exact month it went dark is not something we can verify, so we are not going to invent one.

ReviewMeta is not an isolated case. Fakespot, the largest tool doing the same job, was acquired by Mozilla and then shut down: Mozilla announced in May 2025 that it was winding Fakespot down along with Pocket to concentrate on Firefox itself.

Two things made the category hard to sustain. The analysis depended on scraping review data that Amazon has steadily made harder to collect at scale, and there was never an obvious way to charge for it, because the people who wanted the answer were shoppers making a one-off decision, not businesses with a budget.

Amazon has also moved onto the same ground itself, publishing AI-generated review summaries and reporting enforcement against review abuse. Whether you find that a substitute depends on how comfortable you are with the marketplace grading its own listings.

It depends which job you were using it for, and the honest answer differs sharply between the two.

If you were checking whether someone else's reviews are real, as a shopper or when sizing up a competitor, there is no free like-for-like replacement we can recommend. Several tools claim the capability, but none of the ones we have looked at publish their methodology the way ReviewMeta did, and we are not going to point you at a black box. The manual version still works: sort the reviews by most recent, look for clusters of five-star reviews posted within a few days of each other, check whether reviewers have any other review history, watch for reviews that describe a different product than the one on the page (a sign the listing was repurposed), and compare the rating distribution against similar products in the category.

If you were using it to keep an eye on your own reviews, that is a different market and it is healthy. FeedbackFive, eComEngine, SageMailer and FeedbackWhiz all monitor incoming reviews and ratings on your listings and alert you when something turns negative. Each of them is profiled in this directory.

Nova is not a replacement for ReviewMeta and we are not going to pretend otherwise. It tracks profit, fees and performance across your own catalogue, which is a different question entirely.