Tools and providers
Part of Choosing a reputation supplier after opening the manual evidence first
Why no reputation tool has passed this evidence screen yet
See why this dated England evidence screen admits no named tool, and what an exact product edition must prove before a buyer can compare it fairly.
No named reputation tool has passed this screen yet. Nine were checked against it and each stops at a gate. "Best" is the wording of the search, not a finding. An England buyer cannot compare unnamed editions fairly through generic feature pages, and this review had no authorised hands-on test.
Method: admit only an exact service whose legal entity, UK availability, current edition, contract, data arrangement, support, price basis and exit evidence are documented. Then run the same fictional review-response exercise. Research date: 6 September 2026. Scope: one public-review case for fictional Harbour Lane Furniture Repairs Ltd in England.
Inclusions: software or managed services that can evidence the entire bounded job. Exclusions: directories, awards, testimonials, review totals and category labels. Also demonstrations, incomplete dossiers and other-country editions. Ranking: none. Conflicts and funding: none. Hands-on status: no service was accessed.
The manual baseline uses the buyer's controlled spreadsheet, response document and existing platform account. It is not presumed cheaper, safer or faster.
What to take away
- Nine named services were screened. None passed every gate, so none is recommended or ranked.
- A complete dossier must document legal entity, UK availability, contract, data, support, price and exit.
- The CMA guidance rules out services that depend on fake reviews or inducements.
- The editorial verdict stays HOLD until a full dossier and a common test exist.
Nine named candidates and where each one stops
Each entry is a service a UK buyer can approach, or a free tool they already hold. None clears every gate. The legal entity, the current edition, the price basis or the exit evidence is missing from the public documents.
Google Business Profile. Free, from Google. An owner can reply to Google reviews and report policy breaches. There is no contract, no named support route and limited export, so data and exit evidence stay thin. It suits a sole trader who wants replies only.
Trustpilot. Copenhagen-based, listed on the London Stock Exchange since 2022, with a UK office. A free plan sits beside paid tiers quoted in the low hundreds of pounds a month, typically £200 to £700 depending on tier and billing term. Reviews live on Trustpilot's domain, so ownership at exit needs written terms.
Feefo. Andover, Hampshire. Invitation-only reviews for retail and services. No public list price, so the price basis has to come from a written order form. Complaint handling and reply drafting sit outside the standard service, which leaves the fixture's unsupported allegation uncovered.
Reviews.io. Leicester-based. Monthly plans run in the low hundreds of pounds, typically £99 to £399 depending on invitation volume and feature tier. It covers invitations, collection and replies. Evidence for correction, suppression and deletion of disputed items is not published.
Semrush. Boston-based, listed on Nasdaq. Published list prices as of 2026 are about $139.95 a month for Pro, $249.95 for Guru and $499.95 for Business, lower on annual billing. Its local toolkit handles listings and review replies, not complaint handling or drafting. The UK contracting entity still needs an order form.
Birdeye. Dallas, Texas. Review invitations, chat, payments and surveys in one platform, sold to multi-location groups. No public list price; quotes run per location per month. The contracting entity for an England buyer is not named on the public pricing path.
Podium. Lehi, Utah. Text-message review invitations and messaging for local businesses. Its public pages do not state UK availability or a UK contracting entity, so the buyer would need written confirmation that the exact edition is sold here.
Reputation. San Ramon, California. An enterprise platform sold on annual, quote-based contracts, with published customer work in retail, healthcare and automotive. Public pages give no price basis, no term and no exit terms, so a cost comparison cannot start.
Brand24. Warsaw-based monitoring. Published plans run from roughly $80 to $400 a month depending on mentions tracked. It watches mentions and sentiment only: no invitation, moderation or publication step, so it cannot evidence the whole job.
The CMA's current guidance for businesses and agencies says businesses should not pretend to be customers, commission fake reviews or offer inducements for positive reviews. Any proposition that depends on those practices fails admission.
The CMA's separate guidance for review sites covers collection, checking, presentation, negative material and commercial-interest disclosure. A vendor claim about moderation is not proof that the configured process meets those expectations.
What a candidate dossier must contain
Identity comes before features. Companies House publishes various company-data fields, yet it also states that users remain responsible for applicable data-protection, copyright and other requirements. The buyer must connect the contracting entity to the exact service, order form and support route, treating register entries as limited records.
A complete dossier also needs:
Candidate dossier requirements
- Identityentity, order form, support route
- Monitoring, invitation, moderation, drafting, publishing boundary
- Field-level collection, retention, subprocessors, transfers
- Correction, objection, suppression, deletion evidence
- Authentication, privileged access, logs, incident recovery
- Task-level accessibility evidence, not corporate statement
- Export format, restore check, account ownership at exit
The NCSC's supplier assurance questions cover governance, incidents, data, offshoring and independent assurance. They also cover subcontractors and secure return or deletion at contract end. Supplier answers still need evidence appropriate to the service's risk.
What would change the empty result
Harbour Lane must approve a fictional fixture before any name can appear. It should contain two similar but separate reviewers, a duplicate item, an unsupported allegation, a refused tracking state and an inaccessible task. Then a correction, a deletion request and a failed privileged login. Expected results are written before testing.
Consumer, privacy and PECR, advertising, employment and IP decisions remain independent gates. So do accessibility, security, contract, finance and exit. No average score can conceal a failure. Until a complete dossier and common test exist, naming any of the nine as a pass would be speculation, so the verdict stays HOLD.
The next editorial review should begin with evidence expiry, not another web search for popular names. Any submitted dossier needs a dated owner for every record and a trigger for rechecking changes to terms, subprocessors, functionality or support. An unresolved material change removes the candidate until the same test can be repeated.
Before you act
- Confirm the exact legal entity and UK availability.
- Obtain field-level data collection and retention details.
- Check evidence for correction, objection, suppression and deletion.
- Verify authentication, access logs, incident support and recovery.
- Get task-level accessibility evidence, not a corporate statement.
- Secure export format, restore check and account ownership at exit.
Common questions
Which reputation tools were screened, and why did none pass?
Nine were named: Google Business Profile, Trustpilot, Feefo, Reviews.io, Semrush, Birdeye, Podium, Reputation and Brand24. Each stops on at least one documented gate, usually the exact legal entity, the current UK edition, the price basis or the exit evidence, so none is admitted.
What must a candidate dossier contain?
It needs the precise monitoring, invitation, moderation, drafting and publishing boundary. It also needs field-level collection and retention information, factual data roles, subprocessors and transfers, plus evidence for correction, objection, suppression and deletion handling.
What would change the empty result?
Harbour Lane must approve a fictional fixture before any name can appear. That fixture should include two similar but separate reviewers, a duplicate item, an unsupported allegation, a refused tracking state and an inaccessible task, plus a correction, a deletion request and a failed privileged login.
