Trustpilot’s entire business is selling other companies trust. Its own paperwork reads like the setup to a joke: an active EU fine, a federal lawsuit with its name in the case title, and a Wikipedia page that opens with the word “criticised.” That’s like a smoke detector company whose own office burned down. None of it has cost Trustpilot a single spot on page one for a single “is this company legit” search. Nobody gets that lucky by accident. There’s a specific trick behind it, and there’s a decent chance it’s sitting on your own homepage right now.
The Trustpilot Lawsuit Says the Quiet Part Out Loud
In February 2021, a federal class action was filed against Trustpilot in New York (Trustpilot Damages, LLC v. Trustpilot Inc., case 1:21-cv-00432), on behalf of subscribers described as mainly small and mid-sized companies. The complaint went further than “some reviews are fake.” It said Trustpilot sold subscriptions promising help managing the review system. Then it only delivered that help to a handful of big paying clients. Small and mid-sized subscribers paid the same money and got little for it. It also said that when those smaller subscribers tried to cancel, the renewal emails came from addresses built to land straight in spam. By the time anyone noticed the email, the cancellation window had already closed.
A district court dismissed the case with prejudice in June 2021, and the Second Circuit affirmed that dismissal in June 2022. Legally, the claims didn’t survive scrutiny in either court, and that’s worth stating plainly. What the complaint still put on the public record, regardless of how the case ended, is a specific shape for “they keep asking for money.” Not a subscription. A subscription you can’t quit on schedule.
A cancellation email built to land in spam isn’t a mistake. It’s a very specific kind of customer service.
Trustpilot Fake Reviews: Same Score, Very Different Tuesday
Reviewers on Trustpilot’s own G2 listing describe the same complaint from different angles. Negative reviews vanish for reasons nobody explains, while one-word positive reviews like “Great” or “Amazing” sit untouched for months. One detailed breakdown of the dispute process spells it out: paying businesses get faster escalation when they flag a review, free ones can wait weeks for the same problem. Free customers get customer service. Paying customers get customer service that actually shows up. Even Wikipedia calls out Trustpilot for hosting fake reviews and letting companies delete the bad ones. Wikipedia isn’t exactly trying to start drama.
A PR firm called Alpha Market Flow sells reputation audits to prop trading firms. It ran its own count: 14,277 Trustpilot reviews from 52 firms. Reviews a business specifically asked for, usually sent right after a good outcome, scored 1.03 stars higher on average than reviews people wrote on their own. One-star reviews showed up five times less often in that asked-for group. Alpha Market Flow itself says this doesn’t prove anyone’s cheating. It just shows that asking at the right moment quietly changes who gets counted. It’s their research, and they make money when firms get nervous about this exact number. Trust the finding. Ignore the sales pitch wrapped around it.
A five-star review that lands forty seconds after a payout isn’t feedback. It’s a receipt. Of course he’s happy, you just handed him money.
The Trustpilot Fine: Italy Already Ran the Audit You Were Afraid to Run
In March 2026, Italy’s competition authority, the AGCM, fined Trustpilot four million euros (about $4.6 million) for an unfair commercial practice. The regulator’s own release spells out exactly why. Trustpilot didn’t properly check whether reviews were real, even ones it had stamped “verified.” It also let businesses pick and choose which customers got asked to leave a review. That second part matters more than it sounds: if you only ever invite your happiest customers, the score stops meaning much. The AGCM also found Trustpilot used design tricks that hid which businesses were paying customers, the kind of thing that now has an official name: a dark pattern.
Trustpilot is appealing. It disputes all of it. Here are its own numbers: in Italy, 96% of businesses with at least one review don’t pay for anything. The average score for paying businesses (4.39) is almost the same as free ones (4.40). Both things can be true at once. The regulator’s real problem isn’t the average score. It’s how Trustpilot picked which customers got asked for a review in the first place. Averages don’t show that kind of cherry-picking. You have to check one invitation at a time, which is exactly what the regulator did.
Regulators have a fancier name for it. Everyone else just calls it getting tricked.
What’s a “dark pattern,” exactly?
A dark pattern is an interface built to push people toward a choice they wouldn’t make if it were laid out plainly. Tiny gray “cancel” text next to a giant blue “keep my subscription” button. A checkbox pre-checked to opt you into something. A countdown timer that quietly resets every time you refresh the page. Regulators in the US and EU now treat these as their own category of violation instead of lumping them into ordinary false advertising.
In Trustpilot’s case, the AGCM’s complaint wasn’t about one sneaky button. It was that the site’s design made it hard for an ordinary visitor to tell which businesses were paying customers and which weren’t. That matters because the trust score you’re reading could be shaped by a business relationship you have no way of spotting on the page. Nobody lied to you outright. The truth was just built to be hard to notice.
The “Trustpilot Mafia” Report: Read It With the Right Amount of Salt
Months before the Italian fine, a short seller, Grizzly Research, went further in its own report. It claimed Trustpilot creates review profiles for businesses that never signed up for one. Why? Because a profile nobody claims tends to fill up with one-sided bad reviews. Then, it claimed, Trustpilot sells that same business a subscription to fix the mess it just made. The report says “industry experts” call this the “Trustpilot Mafia.” That’s Grizzly’s own line, not a name anyone else independently confirmed. Trustpilot rejected the report outright, called it factually inaccurate, and countered with its own figures: 97% of businesses on the platform don’t pay for anything, and 70% of the reviews on free profiles are five stars.
A short seller profits directly when a stock drops after a report like this goes out. That doesn’t make the claim false. It does mean the report was never neutral to begin with. File this one under credible enough to note, not settled enough to repeat as fact. Whatever the truth is, Trustpilot’s PR team had a rough week when this dropped.
| What’s alleged | Source | Trustpilot’s response |
|---|---|---|
| Sold subscriptions promising help managing reviews, delivered mainly to large paying clients | Federal class action, SDNY, filed 2021 | Dismissed with prejudice 2021, affirmed on appeal 2022 |
| Two-tier dispute handling, paying businesses escalate faster | WiserReview breakdown of the dispute process | Not directly addressed in public statements found |
| Selective review invitations, “verified” label inadequately checked, obscured which businesses pay | AGCM (Italy), €4M fine, March 2026 | Appealing, cites near-identical average scores paid vs. free |
| Creates unsolicited profiles to farm negative reviews, then sells the fix | Grizzly Research (short seller) report | Categorically rejected, cites 97% of businesses on free tier |
Why Does Trustpilot Still Rank on Google? It Shouldn’t.
Add it up: the EU fine, the lawsuit, the moderation complaints, a short seller’s public accusation. You’d expect at least a modest ranking penalty. Search “[any company] reviews” today and Trustpilot is still sitting on page one, usually above the company’s own website. That’s the actual mystery. “Trustpilot has high domain authority” doesn’t solve it. That’s true, and it explains nothing. Domain authority is an output. Something makes it.
The Trustpilot Widget on Your Homepage Is Doing SEO for Them
I spent ten years rating search quality for Google and Bing. The job trains you to spot this exact pattern: massive trust at the domain level, propping up pages that wouldn’t earn much trust on their own.
Trustpilot’s own marketing to business customers admits part of this. Add the widget, called a TrustBox, to your site, and you get review schema and star ratings showing up in search results. What the marketing copy doesn’t spell out is the other half. TrustBox widgets, like nearly every embeddable trust badge on the web, carry a live link back to trustpilot.com. Trustpilot doesn’t say whether that link actually helps its search rankings. It doesn’t need to. Multiply that one script tag across every business that’s ever installed it, on every page they put it on. It should show up somewhere in the link data.
Pull the actual link profile and the pattern stops being a hunch. Trustpilot and Yelp sit at the identical Domain Rating, 94, the same authority tier by Ahrefs’ own measure. Everything after that number splits hard.
| Site | Domain Rating | Referring domains | Backlinks | Backlinks per referring domain |
|---|---|---|---|---|
| Trustpilot | 94 | 524K | 1.4B | ~2,670 |
| Yelp | 94 | 653K | 201M | ~310 |
| G2 | 91 | 73.7K | 78.9M | ~1,070 |
| Reviews.io | 88 | 21.2K | 20.7M | ~980 |
| Feefo | 83 | 10.9K | 34.3M | ~3,150 |
Yelp actually has more referring domains than Trustpilot. It launched in 2004, only three years before Trustpilot did. That head start barely shows up here. What Yelp doesn’t have is the same habit of getting linked over and over by the same sites. For every single domain that links to Trustpilot at all, that domain links to it roughly 2,670 times on average. Nobody trips down the same flight of stairs 2,670 times. You build an elevator. Then every tenant in the building rides it, and nobody ever asks who owns the shaft.
AI chatbots picked up the same pattern. By the same pull, Trustpilot shows up in ChatGPT’s answers about 414,000 times a month. Yelp matches Trustpilot’s Domain Rating. It even has more referring domains. But it only shows up about 3,300 times, roughly 125 times fewer. Even G2, at 61,000, doesn’t come close, still almost seven times fewer than Trustpilot. Whatever’s driving the backlink gap is driving the citation gap too. It’s the same widget doing both jobs: one script tag quietly telling Google and every AI model that this business vouches for Trustpilot, whether the business meant to say that or not.
Feefo and Reviews.io don’t crack six figures in referring domains between them. Trustpilot could lose ninety percent of its backlink profile in a server fire and still outrank both of them by lunchtime.
That’s the mechanism nobody selling reputation management wants to say out loud. The fix for a bad Trustpilot score is usually more Trustpilot. Every business that takes that fix helps rebuild the same wall around the platform that just got hit with a four million euro fine.
It’s less a trust badge and more a toll booth on your own driveway, for a road that only leads to their house.
How to Read a Trustpilot Score (What Actually Matters)
The star rating alone was never the useful number. Trustpilot labels every review’s origin, invited or organic, and that label sits right on the public profile if you filter for it. Read the organic reviews on their own. Nobody nudged those people into writing them. If a profile has a lot of invited reviews and barely any organic ones, that’s incomplete, not automatically a red flag. And if a business’s Trustpilot score and its Reddit reputation tell two different stories, trust the one that took more effort to write.
This isn’t unique to Trustpilot. Prop firm reviews live and die on exactly this kind of manufactured trust question, which I broke down in more detail in Prop Firm Reviews: SEO, Trust, and Reputation Management. And if you’re trying to work out whether an AI answer engine is quietly repeating a platform’s headline number instead of the real picture, that’s the same mechanism covered in How Prop Firms Get Cited in AI Overviews.
Somewhere right now a business owner is staring at a 2.9-star Trustpilot page, wallet already halfway out, about to pay $299 a month to fix it. Congratulations. You’re about to become referring domain number 524,001.
Got receipts on Trustpilot?
If you’ve got your own story about the subscription pressure, the review gaming, or anything else covered here, come tell me on LinkedIn. I read everything.
Find me on LinkedIn →
