Trustpilot TrustScore Explained: How Its Calculated

What TrustScore actually is

Trustpilot's own support documentation describes TrustScore as an overall measurement of reviewer satisfaction, shown as a number from 1 to 5, based on every Service and Location review a business has received. It's the number that sits next to your star rating on your Trustpilot profile and in any Trustpilot widget embedded on your site — but it isn't computed the way most people assume a "star rating" is computed. A plain average would let a business with five 5-star reviews from three years ago outrank a business with five hundred recent, slightly mixed reviews. Trustpilot's formula is built specifically to avoid that.

How Trustpilot TrustScore is calculated

Per Trustpilot's own TrustScore and star rating explained documentation, the calculation rests on three factors working together:

• Bayesian average. Trustpilot states that it uses a Bayesian average "to ensure that a business with few reviews starts off with a balanced TrustScore," which in practice means the calculation automatically includes the equivalent of seven reviews worth 3.5 stars each in every TrustScore calculation. That synthetic baseline prevents a business with only two or three real reviews from showing a perfect 5.0 or a catastrophic 1.0 on day one — as real reviews accumulate, the baseline's influence on the final number shrinks.
• Time span. Older reviews count for less than newer ones. Trustpilot's documentation is explicit that the formula "gives less weight to old reviews and more to new," so a business that earned strong reviews two years ago but has gone quiet since won't keep coasting on that history indefinitely.
• Frequency. Because the most recent reviews carry the most weight, Trustpilot notes that collecting reviews consistently — as opposed to in occasional bursts — produces a more stable TrustScore over time. A business adding a handful of new reviews every week sees less month-to-month volatility than one that goes quiet for months and then runs a single push for reviews.

Trustpilot also documents how often the number itself refreshes: every new review and rating triggers a recalculation, and for businesses under 10,000 total reviews, that recalculation happens immediately. Above that threshold, TrustScore updates once daily instead of in real time — a detail that mostly matters for very large, high-volume accounts rather than typical small or mid-sized businesses.

What this means in practice

The Bayesian average is the part people most often get wrong. It means a brand-new Trustpilot profile with one glowing 5-star review doesn't show a 5.0 — the synthetic baseline of seven reviews at 3.5 stars pulls the visible number toward the middle until enough real reviews outweigh it. It also means a business shouldn't panic over a single early negative review tanking its score; the same baseline cushions against a tiny sample size swinging wildly in either direction. As total review count grows, each individual review — old or new, good or bad — has proportionally less leverage over the final number, which is exactly what you'd want from a trust metric that's supposed to represent sustained reviewer satisfaction rather than whoever reviewed you most recently.

The time-span weighting is the part that should change your operating habits. Because recent reviews are worth more, a TrustScore built on reviews from 18 months ago is fragile — it's one bad month of service away from a visible drop, since new negative reviews get outsized weight relative to the aging positive ones propping the score up. The fix isn't a one-time review push; it's the frequency factor Trustpilot documents directly, which rewards an ongoing cadence over a campaign.

How to legitimately collect more reviews on Trustpilot

Trustpilot gives businesses purpose-built tools for this rather than leaving it to manual outreach. The platform's automatic invitation methods let a business connect its order or booking system so that a review invitation goes out automatically after a purchase or service is completed, without a staff member choosing who gets asked. Reviews collected this way are labeled "Verified" on the business's profile, which is itself a signal to anyone reading the review that it came through a consistent, systematic process rather than a hand-picked request.

That system-driven approach isn't just more convenient — it's the only approach that's compliant with Trustpilot's own rules. Trustpilot's guidelines for businesses explicitly prohibit being "selective with invitations, by picking and choosing which customers you invite," including "inviting only those people who you know have had a positive experience." An automatic, order-triggered invitation flow structurally can't do that, since it fires for every order rather than every customer a staff member judged likely to leave five stars.

The same guidelines rule out a longer list of tactics worth knowing explicitly, since some of them look like harmless marketing until you read the policy closely:

• Offering "discounts, promo codes, prize draw entries, refunds, freebies, or any other benefit connected with your business for leaving a review" — Trustpilot's guidelines ban incentivized reviews outright, with no exception for disclosed incentives.
• Suggesting "that reviewers only leave positive reviews — for example, 'If you like us, please leave us a 5-star review'" — asking for a positive review, specifically, is treated differently than simply inviting a review.
• Pressuring, blackmailing, or threatening "reviewers to write, edit, or delete a review" after the fact.

Trustpilot enforces this by removing reviews it determines were collected in breach of these guidelines and taking action against the business account — which means a short-term bump from an incentivized review push carries a real risk of reviews being stripped back out later, undoing whatever TrustScore gain the campaign produced in the first place.

Common misreadings of the formula

A few assumptions trip up businesses looking at TrustScore for the first time, and they're worth naming directly since they lead to the wrong fix:

• "My TrustScore should equal my average star rating." It won't, by design. The Bayesian baseline and recency weighting mean TrustScore and a simple arithmetic mean of your stars will diverge, especially for newer profiles or profiles with a lumpy review history — that gap isn't a bug to report to support, it's the formula working as documented.
• "One big review campaign will fix a declining score." A burst of reviews helps, but the frequency factor means Trustpilot's formula specifically favors spreading that same review volume out over time instead of concentrating it. A business better off running automatic invitations continuously than saving up for a single annual push.
• "Deleting bad reviews is the fast way to raise the score." Businesses can't simply delete reviews they don't like — Trustpilot's removal process is tied to its content guidelines (fraudulent, off-topic, or guideline-violating content), not to a business's preference for a higher number. Treat response and service improvement as the lever, not removal requests.

Responding to reviews

Responding to reviews doesn't feed directly into the TrustScore formula the way review volume, recency, and frequency do — Trustpilot's own documentation describes TrustScore as a function of reviewer ratings, not business replies. But responses still matter for the number indirectly: a thoughtful public response to a negative review can prompt a reviewer to update their rating, and a business that visibly engages with feedback is more likely to keep customers willing to leave the next review, which is what actually moves TrustScore over time. Treat responses as part of the review-collection habit, not a separate task bolted onto the side of it.

Comparing TrustScore to how other platforms calculate ratings

Trustpilot's willingness to publish its formula is unusual. Compare that to how Yelp ratings work, where an automated filter decides which reviews even count toward your star rating and the company won't detail the exact criteria, or how Tripadvisor's Popularity Ranking works, which runs on quality, recency, and quantity of reviews but likewise doesn't publish its exact weighting. Trustpilot's documented Bayesian-average-plus-recency approach is the most transparent of the three, and our deeper look at how Google calculates its star rating covers the fourth major platform's own documented mechanics for comparison.

If you're managing reviews across more than just Trustpilot, this is exactly the kind of platform-specific nuance that gets lost when you're checking five different dashboards. Tailwind Score pools your Trustpilot rating together with Google, Yelp, Tripadvisor, the App Store, Google Play, Facebook, and Apple Maps into one 0-100 number, so a TrustScore dip shows up next to what's happening everywhere else instead of sitting in its own disconnected dashboard. Our guide to building an online reputation score walks through how that pooling works across all eight platforms.

TrustScore rewards exactly what good customer service should naturally produce: steady, recent, broad-based reviews rather than a one-time surge. Set up automatic invitations so every customer has a fair, untargeted chance to review you, respond to what comes in, and let the Bayesian average and recency weighting do the rest. If you want to see how that TrustScore fits into your reputation across every platform your customers actually use, try Tailwind Score and track it all from one place.