The short answer

Treat reviews as one evidence source, not a verdict. The star average compresses many unknowns: who wrote the reviews, whether they used the exact product, whether they received an incentive, how recent the experience is, and what the platform chose to display.

The goal is not to announce that a review is fake. It is to decide whether the available evidence is strong enough to influence your next step.

Use the five-question evidence check

1. Who controls the source?

Is the review on a retailer, manufacturer, independent publication, creator channel, or company-controlled comparison site? The FTC recommends considering the source and using a variety of sources, including trustworthy independent expert reviews.

2. How recent is the evidence?

Check whether the review describes the current product and version. The FTC also says a sudden burst of reviews can sometimes warrant additional checking. Timing alone does not prove manipulation.

3. Is reviewer history visible?

When the platform permits it, examine whether the account has other relevant reviews and whether the experience described matches the exact product. A single-review account can be a reason to ask questions, not proof that the review is false.

4. Is an incentive or relationship disclosed?

Look for labels describing a free product, sponsorship, employment, affiliate relationship, or other material connection. Missing information is an unknown; do not invent a relationship.

The FTC's current Reviews and Testimonials Rule guidance also distinguishes consumer reviews from promotional testimonials and warns about misrepresented use or experience.

5. Do other evidence families agree?

Compare the exact manufacturer/model details, official instructions, professional testing methods, independently disclosed limitations, and any later hands-on evidence. The FTC's online-shopping guidance says not to rely on star ratings alone and recommends reviewing multiple sources.

Use evidence-state labels

Digital Derek can classify a claim as:

  • official claim: what the maker or platform states;
  • professional test: a disclosed method and observed result;
  • disclosed incentive: experience with a visible material connection;
  • customer report: a self-reported experience that may be useful but is not independently verified;
  • unknown: missing source, method, relationship, version, or context.

These labels describe the evidence available. They do not declare that a person is truthful, dishonest, real, or fake.

What not to do

Do not copy customer reviews into a new score, run private or restricted review data through an unapproved scraper, accuse a named person or seller, or treat a writing style as proof of AI generation or fraud.

The FTC explicitly warns that it can be nearly impossible to distinguish real and fake reviews just by looking. A checklist should produce `stronger`, `weaker`, or `unknown` evidence—not a fake-review conviction.

What you still need to check

  • Warning signs cannot prove that an individual review is real or fake.
  • Reviews cannot confirm how the exact model will perform for your use; compare product details and independent evidence too.
  • If you suspect fraud or a rule violation, use the platform's reporting tools or official consumer channels. This checklist is not a legal judgment.

Bottom line

Use more than the stars. Ask who controls the source, how current it is, what is known about the reviewer and incentives, and whether independent evidence agrees. Preserve `unknown` when the available information cannot answer.

See how Digital Derek separates research from a tested verdict.