Measurement
Reputation management measurement in review software
Reputation management measurement in review software: configure records, measures and checks that support CMA fake-review controls and reporting.
What to take away
- Overall ratings, summaries, review counts and rankings are consumer review information, and publishing them in a misleading way is banned.
- Publishers must take reasonable and proportionate steps to prevent and remove banned reviews, which is a positive obligation.
- The general steps range from risk assessment to regular effectiveness checks; map each one to a software record you can review.
- Paragraph 13 of Schedule 20 to the Digital Markets, Competition and Consumers Act 2024 has been in force since 6 April 2025.
- Reply coverage and timing are separate internal measures, covered in the key metrics article.
Which numbers are regulated
Reputation management measurement in review software starts by separating figures that describe what customers wrote from figures that describe your service. The first kind falls under consumer protection law, so it needs more care.
Paragraph 13 of Schedule 20 to the DMCC Act 2024 bans publishing consumer reviews, or consumer review information, in a misleading way. The Schedule 20 banned practices include examples, and the definition is wide.
In the CMA's fake reviews guidance (CMA208), examples of consumer review information include aggregated information in the form of overall ratings, overall summaries, review counts and rankings. Because of the value of this information to consumers' decision making, it is important it is accurate.
Examples of misleading publication
Publishing in a misleading way includes, for example, failing to publish, or removing from publication, negative consumer reviews whilst publishing positive ones, or vice versa. Giving greater prominence to positive consumer reviews over negative ones is another example.
Omitting information relevant to the circumstances in which a review was written is a third. That includes the fact that a person was commissioned to write it.
A fake consumer review purports to be, but is not, based on a person's genuine experience. Keep that definition next to your measurement mistakes list, and read fake Google reviews for removal routes.
The six steps and what each one produces
When setting up review or reputation-management software, map the records below to available fields, logs and case notes so the steps can be checked from your system. The obligation is positive. It requires traders who publish consumer reviews or consumer review information by any means to take effective action, rather than merely refraining from a specified action. The precise steps depend on the circumstances, but the guidance sets out general steps likely to be reasonable.
| Step in the guidance | Record to keep | Illustrative measure |
|---|---|---|
| Assess the risk of banned reviews | Dated risk assessment | Date of last review |
| Publish a policy, including incentivised reviews | Policy version and date | Version number |
| Detect banned reviews | Flag log with source | Reviews flagged per month |
| Investigate | Case notes and outcome | Cases open and age in days |
| Act, including sanctions | Removal or reporting record | Reviews removed per month |
| Assess effectiveness regularly | Dated effectiveness review | Date of last review |
Example: how selection changes an average
This is illustrative arithmetic. A branch has 100 reviews: 90 five-star reviews and 10 one-star reviews.
| Basis | Calculation | Average | Count shown |
|---|---|---|---|
| All reviews | (90 times 5 + 10 times 1) divided by 100 | 4.6 | 100 |
| One-star reviews left out | (90 times 5) divided by 90 | 5.0 | 90 |
The second line raises the average by 0.4 and cuts the count by ten. Genuine one-star reviews removed only because they are negative fall under the examples above. A removal log with a reason for each case is your protection.
Incentivised reviews and the disclosure test
Traders are free to incentivise reviews, but to comply with the law they must tell consumers that the review has been incentivised, and the review must still reflect the reviewer's genuine experience.
Add a field to the review register for incentive status. It lets you exclude or label those reviews in any average you publish, and it supports the published policy. Read multi-location review monitoring for a register that covers many branches.
Data protection and reporting limits
Review records usually contain names and sometimes other personal information. You must identify and document a lawful basis under the UK GDPR before using personal information. The Data Protection Act 2018 also applies; check Information Commissioner's Office (ICO) guidance when setting access and retention rules. See data protection for roles and retention.
Show the limits on the face of the report: the period, the extraction cut-off, the branches covered and any missing timestamps. A reporting dashboard should carry those labels beside each figure. For response speed, see the review response time benchmarks.
Checklist
- Date the risk assessment and the policy version.
- Log every flagged review with its source and outcome.
- Record why any review was removed and who approved it.
- Reconcile the review register to the published count each month.
- Label star averages with the date and the number of reviews behind them.
- Hold an effectiveness review at least once a year and record it.
- Document the lawful basis for review data.
Common questions
Is a star average on my website regulated?
An overall rating is consumer review information, so publishing it in a misleading way is banned. Keep the register of reviews behind it and reconcile the count.
Can I offer a discount for a review?
Traders may incentivise reviews, but they must tell consumers the review has been incentivised, and it must still reflect the reviewer's genuine experience. Check the platform's own rules as well.
Does the rule apply to reviews on other platforms?
It applies to traders who display or make available consumer reviews or consumer review information by any means, including on their own media.
In this guide
- Reputation metrics with review cohorts and reply clocks defined properlyReputation management key metrics for England businesses, with review cohorts, reply timing, rating definitions and clear limits on outcome interpretation.
- Reputation management reporting dashboard: reconciliation firstReputation management reporting dashboard design for England teams, covering source reconciliation, missing data, response queues and export checks.
- Reputation management attribution methods: credit vs impactReputation management attribution methods separate credit for recorded paths from estimates of impact, using examples, review records and UK review rules.
- Five reputation measurement mistakes, starting with the displayed ratingReputation management measurement mistakes explained through a non-ranked checklist, documented selection method, source records and practical corrections.
- What the UK fake-review research can and cannot tell your reputation teamReputation management benchmark research for England teams: interpret dated UK fake-review evidence, check comparable units and reject unsuitable targets.
