How to spot fake reviews
Review fraud is an industry. Sellers buy positive reviews for themselves and occasionally negative ones for competitors, and the people writing them are paid by volume. Volume leaves a signature.
Look at the shape of the timeline
Genuine reviews arrive irregularly, in step with how many people use the service. Forty five-star reviews inside one week, then silence for three months, is not how that works. A cluster right after a launch, or right after a wave of complaints, deserves suspicion.
Genuine reviews are specific and slightly boring
Someone who actually used a service mentions the plan they were on, how long delivery took, the name of the feature that broke, what support said. Paid reviews stay general because the writer has no details to give: “great service, highly recommend, will use again” could describe a dentist or a hosting company.
Perfect scores are a warning, not a recommendation
Any service with real customers has unhappy ones. A profile with hundreds of reviews and no criticism has been curated, either by buying praise or by removing complaints. A four-point-something with visible, answered complaints is usually more trustworthy than a flawless five.
Read the negative reviews first
One angry review says little. Six describing the same failure, months apart, describe the service. Also read how the owner responded: a specific reply that addresses the actual problem tells you more about a company than any amount of praise.
Check what the platform requires
Ask what it takes to post a review. Can one person post ten? Is an account required? Can the subject of the review pay to have it taken down? On Phyco an account can leave one review per website, the email address has to be confirmed, and average scores only appear once there are at least three reviews. We do not write reviews ourselves and we do not remove them on request. The rules are in the terms and the reasoning is on the about page.
Signals worth trusting
- Concrete details: dates, plan names, amounts, error messages.
- Mixed sentiment inside a single review: what worked and what did not.
- Reviews spread over time rather than arriving in bursts.
- Owner replies that answer the complaint instead of thanking the reviewer.
- A platform that is explicit about who may review and what it removes.