A rush of reviews in a few days
When a batch of reviews lands within about a week and little is known about the people who left them, we flag that cluster for a closer look before treating it as strong proof.
How we check the reviews we show.
We publish reviews from named public sources, not reviews we collected ourselves. Every review is shown with its source named, and we run the same checks across all of them to spot the patterns that fake reviews tend to leave behind. These checks matter twice: they shape what you read here, and what AI assistants read when you ask them about a trader using our site.
What we look at
The patterns we check on every trader.
A jump above a business's own normal pace
We compare the last week of reviews against that same business's recent history, not a generic average, so a genuinely busy spell reads differently from a sudden unexplained surge.
The same review text on more than one site
Near-identical wording that turns up across different platforms stands out, because real customers rarely post the same words in two places.
How much a source tells us about its reviewers
Where a site normally shows who left a review, we take that into account. A site that never publishes reviewer names is not treated as suspicious for that alone.
Whether different sites broadly agree
We put every rating on the same scale before comparing, so a 9 out of 10 and a 4.5 out of 5 line up. A wide gap that remains becomes a caution to check, not a claim that anything is fake.
Whether the headline count matches the sources
We show how many full reviews we display and how many are counted across public sources, so the headline number lines up with what those sources actually publish.
When something looks off
What we do when a pattern stands out.
We show a caution, in plain words
The profile carries a short, readable note about what looked off, so you can weigh it yourself. A caution flags a pattern to check; it is not a ruling that reviews are fake.
We keep the history
We do not delete or hide a business's review history. The reviews stay visible with their source named, right next to the caution.
We lean towards not flagging
When a pattern has an innocent explanation, such as a review site that never publishes names, or an old surge that is no longer recent, we hold back rather than raise a caution the business has not earned.
What we cannot know
Where these checks stop.
- We cannot confirm from a review alone that a job actually happened.
- A clean set of checks is not a guarantee of good work.
- We are not the review sites. We show what they publish, with each source named, and link back to them.
- A caution marks a pattern worth a closer look, not proof that anyone did anything wrong.
Corrections
If we get it wrong, we fix it.
A trader who thinks a caution or a fact is wrong can tell us through Profile rights. We look into it and put genuine errors right.
- Businesses can challenge any caution, or correct a factual error, through Profile rights.
- We act on evidence, never on payment.
- We investigate every report and reply within 48 hours.
- When the evidence changes, the caution changes with it.
Check a trader, then read the standard.
Use the checks to compare quickly, then verify the evidence before you hire.