Media buying agencies launch review: a desk protocol

A media buying agencies launch review should answer one narrow question: did the approved campaign reach live systems without a material control failing? Review the first-party account record against the signed brief and build sheet. Do not judge an agency from a polished launch report alone.

What to take away

Review method and limits

This page sets out a desk-review protocol. It does not report hands-on testing, purchases, interviews or campaign results. No agency or platform has supplied a sample, sponsorship or affiliate payment. This is a general method, not an account-specific review: before a real review begins, the account owner must assign a named reviewer responsible for checking campaign controls and record that person's name, responsibility, review time and conflicts.

The reviewer should receive the approved brief, media plan, creative sign-offs, build sheet, live account exports, consent decision, measurement test and change log. Request a dated export or audit record from the platform in use. Examples include Google Ads Change history, Meta Ads Manager Activity history and Display & Video 360 Change history; retain the relevant account and campaign records, including who changed what and when where available. Platform screenshots are useful, but a dated export or audit record is easier to preserve and compare. Product interfaces and fields can change, so the platform's own record must be read in its current context.

Trace the launch, do not rescore the strategy

Start with five samples: one campaign, one audience, one creative item, one destination and one measurement event. Trace each from approval to the live setting. Expand to all live items in the affected sample class if any discrepancy appears, or when the campaign contains materially different formats. Treat a mismatch in an approved date, geography, budget control, exclusion, consent state, destination or measurement event as control-relevant; do not apply a percentage tolerance to these fields.

Check dates, England or wider UK geography, budget controls, exclusions, status and account identity. A live item outside the approval boundary is a launch defect even if early results look favourable. Record the exact mismatch, owner and containment action.

Review privacy and content gates

For UK online advertising, the Information Commissioner's Office (ICO) Guidance on the use of storage and access technologies explains the rules for storing information on, or accessing information stored on, a user's device. Regulation 6 of the Privacy and Electronic Communications Regulations 2003 (PECR) generally requires consent for those activities unless an exemption applies; where personal data is processed, the UK GDPR also applies. The launch reviewer should check the recorded consent decision and confirm that a qualified person approved the actual setup. This desk protocol cannot make a compliance determination.

For influencer executions, the UK ASA and CAP resource hub points brands, agencies and creators to current disclosure and social advertising material. Check the released creative against the approved disclosure treatment and the applicable CAP Code rules, including rule 2.1 on making marketing communications obviously identifiable and rule 2.4 on identifying advertorials as marketing communications. Apply the relevant rule to the format.

Campaigns using reviews or endorsements need an origin and incentive check. The UK Competition and Markets Authority collection addresses brands, media agencies, creators, platforms and review publishers. It supports the review of disclosure and authenticity controls, not an opinion about the agency's overall competence.

Inspect early delivery carefully

Confirm that spend has begun only where intended, approved assets are serving and agreed events are arriving. Early totals are diagnostic, not proof of effectiveness. Keep platform-side delivery separate from the client's commercial outcomes, and label incomplete data.

Where the reviewer finds a serious mismatch, pause the affected element and preserve the evidence. A small labelling error may need a creative correction; an audience, consent or uncontrolled-spend problem may require wider containment. The authorised owner decides the extent.

Record the launch decision

Use pass, conditional pass or stop. A conditional pass needs a named action, deadline and acceptance owner. The final note should list the artefacts inspected, sample limits, defects, corrections and unresolved points. That creates an auditable launch record without pretending that a short initial review predicts campaign performance.

Schedule a follow-up around the risks found rather than a standard ceremonial meeting. Recheck corrected items directly in the source system and compare later exports with the launch snapshot. If a defect recurs, examine the handover and checking method instead of recording the same isolated fix again.

Before you act

Common questions

What is the narrow question a media buying agency launch review should answer?

It should answer whether the approved campaign reached live systems without a material control failing. The review compares the first-party account record against the signed brief and build sheet. It does not judge an agency from a polished launch report alone.

How should a reviewer trace a campaign launch?

Start with five samples: one campaign, one audience, one creative item, one destination and one measurement event. Trace each from approval to the live setting. Expand the sample when a discrepancy appears or when the campaign contains materially different formats.

What should a reviewer do when finding a serious mismatch?

Pause the affected element and preserve the evidence. A small labelling error may need a creative correction. An audience, consent or uncontrolled-spend problem may require wider containment. The authorised owner decides the extent.