Operations
Part of Reputation management operations: from source record to publication
Reputation-management quality checklist for England: CMA and Google controls
Check review software workflows in England against named CMA guidance and Google policies, with evidence records, disclosure checks and release decisions.
The case starts with an unresolved review
The worked case starts with one review whose rating is challenged, while the available record does not establish whether it is authentic. Its decision stays open. Disagreement with the rating alone is not a reason to delay or suppress a genuine negative review before publication.
Treat the individual review record as the control unit. Keep its source item, evidence checks and decision in reputation-management software or a case tracker; automated flags direct attention but do not replace the responsible reviewer. Keep supplier screening separate from publication checks, using a market entry checklist before review decisions.
At intake, test the configured workflow rather than confirming only that a field or flag exists. Record what was examined, who made the decision and what remains unresolved for this review. A blank field does not explain the evidence or the decision, so it cannot close the case.
Applicable rules are identified before the evidence is weighed
The reviewer first maps the legal and platform controls to the work the software supports; that mapping does not settle whether this review is genuine. Review and endorsement rules apply to people who create, commission or publish reviews or endorsements, and to those who facilitate those activities.
The Digital Markets, Competition and Consumers Act 2024 introduced a banned practice covering fake reviews and concealed incentivised reviews. Concealment includes hiding an exchange for money or another benefit. Conduct within a banned practice is automatically unfair and illegal; the statutory schedule sets out the relevant controls.
Reviews and online content must be genuine and accurate, while endorsements must be clear. For business websites that accept product reviews, the CMA online review site guidance says genuine, relevant and lawful reviews should all be published and processes must not obstruct that. The reviewer records which activities the workflow supports.
The source is kept separate from its rating
The source check preserves the original submission and tests it against platform requirements without treating sentiment as proof. Google Maps User Generated Content Policy says a review or rating should reflect an actual business experience, be genuine and unbiased, and that fake engagement is prohibited and will be removed.
The finding in this case is still unresolved: the original submission is retained, but the available checks have not established authenticity. Neither the rating, favourable tone, automated label nor sentiment score proves an actual experience. Record evidence for the decision and make missing evidence visible rather than converting uncertainty into an authenticity conclusion.
Collection and display are checked as separate controls
The reviewer tests how reviews are invited, examined and presented, applying appropriate rigour to both positive and negative items. Authenticity, relevance, rating treatment and prohibited-content questions stay distinct, and every decision is tied to evidence in this review record rather than collapsed into one automated flag.
| Control checked | Required handling |
|---|---|
| Scrutiny | Examine positive and negative reviews with appropriate rigour, use procedures to prevent and remove fake reviews, and record evidence for each decision. An automated label, sentiment score or favourable tone is not proof of a genuine experience. |
| Incentives and selection | Do not offer payment, discounts, free goods or services in exchange for posting, revising or removing a review, including a negative review. Do not discourage or prohibit negative reviews or selectively solicit positive ones. Inspect invitation wording and selection rules. |
| Genuine negative reviews | Do not edit, withhold, remove or delay genuine negative reviews. Criticism is not a suppression reason. Record any separate authenticity or content concern, the evidence examined and the decision. |
| Relevance and ratings | Reviews should relate to the product described, not be outdated, and be published without unreasonable delays. Aggregate ratings should include all genuine reviews without distortion from fake and incentivised reviews. Tie rating decisions to the review and moderation record. |
| A clear process | Publish a policy prohibiting fake reviews and explaining the approach to banned content. Explain how reviews are collected and checked, what users need to do for publication, and why a review might not be published. Clearly and prominently disclose commercial relationships with listed businesses. |
The record assigns an owner rather than guessing
With authenticity unresolved, the reviewer makes no unsupported conclusion and assigns the open question to an owner with a next action before approval. Moderation remains separate from release checks: a negative rating cannot justify holding a genuine review or waiting for a more favourable outcome.
Keep the original review apart from internal notes, and separate source capture, evidence review, draft response and publication decisions in the operating workflow. Corrections should attach to the original decision rather than silently replace its history. Once a response or policy change is live, compare it with authorised wording and record any difference.
Record incentives and commercial relationships in a disclosure policy and assign an owner to compare public wording with the recorded relationship. UK data protection is governed by UK GDPR and the Data Protection Act 2018. The accuracy principle requires personal details to stay accurate and up to date; a data protection workflow routes disputed personal information separately from authenticity decisions.
Build a reporting dashboard around traceable case outcomes rather than unexplained totals. Retain underlying evidence and each decision, and give every open issue an owner and next action. Compare the checks with a review methodology; neither a blank field nor a favourable overall rating proves every control has passed.
Rehearse risk scenarios through fictional records, including an unsupported reply, missing evidence and an unapproved publication change. For this case, that exercise should keep the original submission, moderation reason, evidence gaps and release choice traceable without turning a rehearsal into a finding about authenticity.
Release follows the recorded result
Release depends on the evidence and assigned status, not on a completed form alone. A pass retains evidence and approval, a fail returns the draft or policy change for correction before release, and an unresolved result identifies missing evidence or a decision and assigns an owner.
Result / Next action
- Pass
- Retain the evidence and approval with the case record.
- Fail
- Return the draft or change for correction before release.
- Unresolved
- Identify the missing evidence or decision and assign an owner.
Rules carried into later cases
The next case follows the same discipline: test the source, separate authenticity, disclosure and response wording, record evidence, and keep unresolved matters assigned rather than implied to be settled. A completed checklist is not itself a legal conclusion or proof that controls passed.
Before publication, resolve each question against the source review, intended action and supporting evidence. Ask the responsible reviewer to give a reason for every action; a general approval is insufficient if an individual concern remains unexplained. Keep authenticity, disclosure and response wording distinct when recording the decision.
If only the rating is disputed, distinguish that disagreement from evidence about authenticity, relevance or prohibited content. Apply appropriate scrutiny to positive and negative reviews. For an aggregate rating, check that all genuine reviews are represented without distortion from fake or incentivised reviews.
