Fictional example: 4.8 stars, two different issues

An invented subscription service shows 4.8 out of five from 120 ratings. Its highlighted review says ‘Five stars: easy to set up’. An older one-star review says ‘The evening reminder arrived late; I also could not find how to cancel’. No real service or users are described. The average does not say that 96% quit, or that 120 people stopped smoking.

The praise is about setup. The complaint raises reminder delivery and cancellation as two separate product questions. Note the older review's version and market. A current help page might explain how to cancel without telling you whether the earlier user was refunded; a release note about reminders might describe a repair without showing it works on every device.

You can ask the provider: ‘Where are the current cancellation conditions for this market?’ and ‘Which version changed reminder delivery, and what limitations remain?’ Those enquiries need no smoking history. If neither answer is available, leave the issue unresolved instead of using the overall stars to cancel out the complaint.

Who contributed to the average?

Look for how reviews are requested, filtered and displayed, whether all ratings enter the average, and whether gifts or discounts are disclosed. The FTC's U.S. platform guide distinguishes collection, moderation and publication, warns about selective positive invitations and says even buyer-only systems are not immune to fake reviews. These are reading questions, not a legal verdict about this example or another country.

Switching from highlighted to recent reviews can reveal different issues if that option exists. Repeated wording or a sudden cluster warrants a question about authenticity; it does not prove each reviewer is fake. A purchase badge concerns the transaction under that platform's rules, not verification of abstinence. Nor are the 120 reviewers necessarily all customers.

[1]

Genuine satisfaction still leaves the outcome question open

NCCIH distinguishes anecdotes and testimonials from objective research information. To support a claim that a service changes quitting outcomes, you would need research defining the participants, what they received, a meaningful comparison, the outcome and follow-up. A rating page does not supply those fields merely by showing more stars.

The same limit applies to low ratings: a billing problem may be real without establishing that the service has no quitting effect. Keep the operational question precise, and read outcome evidence separately. This page neither ranks services nor evaluates an intervention.

[2]

Check the product question without exposing a private account

Use current public help, terms and release notes to check language availability, accessibility, charges or a reported fault. A provider's statement can answer what it currently offers; it is not independent proof of a health outcome. Do not post receipts, account screens or smoking records to win a debate about a review.

Support or treatment decisions for yourself belong with a qualified local health professional or cessation service. Reviews can help you formulate a practical question, while the professional discussion addresses suitability. For a transaction dispute, use the platform or local consumer-help route separately.

What to keep in mind

  • Stars summarize feedback, not a defined quitting outcome.
  • A reminder fault and cancellation terms need separate answers.
  • Authenticity, current product information and outcome evidence are different checks.

Sources

The central claims on this page were checked against the sources below.

  1. U.S. Federal Trade Commission: Featuring Online Customer Reviews: A Guide for Platforms

    Sources checked: 2026-10-04

  2. NCCIH — National Institutes of Health: How Do You Know the Information Is Accurate?

    Sources checked: 2026-10-04

General review-reading guidance. Ratings, service and reviews in the example are fictional. No product recommendation, health forecast or legal conclusion is made.