
Rules and ethics
Part of Reputation management UK rules: ethics and software checks
Reputation management commercial contracts: writing the schedule
Reputation management commercial contracts: write the service schedule with a line per job, sample wording, acceptance tests, change notice and pause rights.
A reputation management contract can shrink to one promise: to improve reputation. The service schedule is where that promise has to become separate jobs, each with an owner, a test and a rule for change. Without it, a buyer cannot say what was bought, what was delivered or who may stop the service.
The schedule comes after the supplier has been vetted, which the evidence room covers. Counsel should settle the final wording, but the lines below give the schedule its shape.
What to take away
- Give collection, monitoring, moderation, complaint handling and response writing separate lines in the schedule.
- Name the legal entities, edition and territory in the signed schedule.
- Define deliverables, evidence sources and acceptance tests without inventing thresholds.
- Require notice of changes, and name who may pause ingestion or publication.
- Signature stays on hold until every gate has evidence.
Give each job its own line
Collection, monitoring, moderation, complaint handling and response writing should each appear as a separate line in the schedule. Buried inside a promise to improve reputation, they cannot be tested or owned. A separate line also shows which party acts at each point in the review chain.
That split matters in law as well as in practice. Government guidance from the Competition and Markets Authority (CMA) says the law on reviews and endorsements applies to you if you create, commission or publish reviews or endorsements, or facilitate these activities. The Digital Markets, Competition and Consumers Act 2024 is the statute behind it.
The schedule should say which party does each of those things.
| Job line | What the schedule should name | Evidence to request |
|---|---|---|
| Collection | Sources, who invites reviewers, incentive records | Sample of source records |
| Monitoring | Platforms watched, such as Google reviews or Trustpilot, and who checks status | Monitoring log sample |
| Moderation | Who decides what stays, authenticity checks | Moderation policy |
| Complaint handling | Who receives complaints, how they stay apart from reviews | Complaint separation log sample |
| Response writing | Who drafts, who approves, who publishes | Approved response record |
Here is how one line might read in the schedule: "The Supplier moderates reviews for the Customer. The Customer's named reviewer approves any removal request before it is sent. The Supplier keeps a log of every decision and gives the Customer access to it on request." It names the actor, the approval and the evidence, so the line can be tested.
Record incentives in a disclosure policy, so the collection line has evidence behind it when someone asks how a review came to exist.
Name the parties, edition and territory
The signed schedule must name the supplier's legal entity, the buyer entity, the edition bought and the territory covered. A trading name is not enough. Attach the schedule to the order form, so the edition in scope can be proved later rather than recalled.
Evidence to request: a signed service schedule with legal names, edition and territory. Check it against the order form before signature, and record any difference as unresolved, because a mismatch between the two documents is a contract risk.
Define deliverables and acceptance
Deliverables, evidence sources, versioning and correction times belong in the schedule, and none of them needs an invented threshold. Say what is delivered, from which source, in which version, and how a fault is corrected. Leave numeric targets until the buyer has measured its own baseline.
Acceptance tests turn each line into a pass or a fail. A record of each test should carry four things:
- The input used, such as a real review case with personal data removed, since UK GDPR applies to any personal data left in.
- The result the schedule expects.
- The result observed, with its evidence.
- The outcome, owner and expirypass, fail or unresolved.
The selection checklist sets out acceptance exercises for the buying decision. Keep the test records with the contract, so a later dispute about delivery starts from evidence and not from memory.
Control change after signature
Require notice of changes to moderation, integrations, automation and support. Each of those can alter what a published review looks like, so the schedule should give the buyer time to retest before the change goes live. Record who receives the notice and who approves it.
State who may pause data ingestion or publication when an independent gate fails. A pause right that needs the supplier's agreement is not a control. Name the buyer role that can act without waiting, and the evidence the supplier must give afterwards.
Currency, VAT display, billing unit, renewal, additions, suspension, liability, termination and transition work should also appear in the schedule. Price any later addition as its own line, so a change in scope is a visible decision.
Say who handles complaints, appeals and corrections
Allocate responsibility for negative reviews, platform appeals and corrections in the schedule. For website operators, the Ministry of Justice's section 5 guidance relates to a defined England and Wales procedure. It is not a generic deletion promise, so the schedule should not offer one.
The procedure comes from the Defamation Act 2013 and the Defamation (Operators of Websites) Regulations 2013. Counsel must assess whether that procedure applies to the buyer's site at all. Where it does, the schedule can name who forwards a notice, who decides and who records the outcome. Where it does not, say how a complaint reaches the platform instead.
Keep signature on hold until every gate has evidence
Assign every line an evidence owner, an expiry and a pass, fail or unresolved outcome. Do not average privacy, review integrity, security, accessibility, IP or exit against the schedule. Signature stays on hold when any necessary gate lacks evidence.
Before you act
- Give each job its own line in the schedule.
- Name the legal entities, edition and territory.
- Write an acceptance test record for every line.
- Require notice of changes to moderation, integrations, automation and support.
- Name the buyer role that may pause ingestion or publication.
- Allocate responsibility for complaints, platform appeals and corrections.
Common questions
Why split the service into separate lines?
Collection, monitoring, moderation, complaint handling and response writing carry different risks and owners. Separate lines let the buyer test, accept and, if needed, pause each one without stopping the others.
How should each line be tracked?
Give each line an owner, an expiry and an outcome of pass, fail or unresolved. Do not average the risk areas.
Who should be able to pause the service?
The schedule should name a buyer role that can pause data ingestion or publication when an independent gate fails, without waiting for the supplier's agreement.



