Six influencer agency service measures and how to define each one

Influencer agencies service standards should define how evidence moves, not claim a universal turnaround. The buyer sets thresholds after examining campaign risk, workload and baseline data. Until then, keep the threshold field blank. Thresholds are buyer-owned and blank by design, so no supplier number can fill them in for you.

What to take away

Write a complete measure card

Every card needs:

Worked example card (illustrative figures)

In England, baseline data comes from three places: agency reporting exports, signed contract terms, and prior campaign logs. Agency reporting supplies the numerator and denominator as the agency recorded them. Contract terms fix the calendar, timezone and pause rules. Prior campaign logs show the spread of past results the buyer can compare against.

Brief evidence completeness

Population: briefs submitted for formal intake in the reporting period. Numerator: briefs containing every buyer-defined mandatory field and attached evidence. Denominator: all eligible submitted briefs. Exclude test records and documented duplicates. The clock is irrelevant unless the buyer also measures return time separately.

Evidence comes from the versioned intake register. Failure response: return the brief to the advertiser owner and keep creator search on hold. CAP Code claims rules support treating substantiation as a distinct input, but do not supply a performance threshold. In practice, file the substantiation record with the brief version so a reviewer can trace it.

Final-asset approval lead time

Event: receipt of a complete, versioned final asset. Unit: elapsed working time under the buyer's calendar and timezone. Start only when required files and evidence arrive. Pause for a logged buyer dependency; stop at the final mandatory approval or rejection. Report the distribution and exceptions, not only a mean.

CAP's influencer recognition guidance makes presentation context relevant. The evidence should therefore record which final format and preview reviewers opened. In practice, log the label text and the preview link beside the version number.

Controlled-release fidelity

Population: published items authorised in the period. Numerator: items whose initial live capture matches the approved version, account, link, disclosure and media instruction. Denominator: every eligible release. A later edit is a new version, not an automatic pass.

The NCSC's social-media account guidance supports controlled access and recovery planning. The service record must still identify the actual publisher and account event. In practice, name the account holder and the recovery contact for each release.

Correction handling time

Event: a validated incident requiring correction or withdrawal. Unit: elapsed time from triage acceptance to the authorised containment action. Define who can validate, which severity classes are included, when the clock pauses and what counts as containment. Do not merge investigation and final remedy unless the contract says so.

For a personal-data event, follow the separate privacy incident route. The ICO's breach guide explains UK assessment and notification duties for relevant breaches; the service clock cannot replace them. For an agency handling creator and audience data, those duties come from UK GDPR and the Data Protection Act 2018, and the ICO is the regulator. In practice, keep the separate privacy log and cite it in the closure pack.

Creator payment reconciliation

Population: creator milestones accepted during the period. Numerator: milestones with consideration, invoice or payment evidence reconciled under the contract. Denominator: all eligible accepted milestones. State how disputes, missing bank details, credits and non-cash consideration are handled. A prompt payment measure says nothing about creative quality. Where the contract is silent on late payment between businesses, the Late Payment of Commercial Debts (Interest) Act 1998 may give the creator a claim to statutory interest, so log the due date for every milestone.

Closure evidence completeness

Population: campaigns presented for closure. Numerator: campaigns with resolved corrections, payment records, rights status, media shut-off, access revocation, export and retention or deletion decisions. Denominator: all closure candidates. Any mandatory unknown blocks closure rather than lowering the result silently.

Reviewers should inspect source records, excluded cases and query logic on a buyer-owned schedule. If a threshold encourages premature approval or incident misclassification, suspend it and repair the definition before judging the service.

Before you act

Common questions

What must every measure card contain?

Each card needs the event and eligible population, plus a numerator and denominator or a clear unit. It also needs clock start, pauses, stop and timezone. Add inclusions, exclusions and missing-data treatment, the source system and report version, an owner and review frequency, and a blank buyer-set threshold.

How should final-asset approval lead time be measured?

Start the clock only when required files and evidence arrive, using elapsed working time under the buyer's calendar and timezone. Pause for a logged buyer dependency and stop at the final mandatory approval or rejection. Report the distribution and exceptions, not only a mean.

What counts as a pass in controlled-release fidelity?

The population is published items authorised in the period. The numerator is items whose initial live capture matches the approved version, account, link, disclosure and media instruction. The denominator is every eligible release. A later edit is a new version, not an automatic pass.