Nine checks in an out of home advertising launch review

What to take away

• Most launch reviews fail because they assess the creative and the site list, then leave audience measurement, data protection and ownership of post-campaign numbers unexamined.
• This is a desk review method: it separates what a vendor asserts from what independent evidence shows, without hands-on testing.
• Nine checks cover the brief, the sites, the rules, the data and the reporting line before money is committed.
• A launch review is not a sign-off ritual. It is where someone is named as accountable for each claim in the plan.
• Where a claim cannot be sourced, the review records it as unverified rather than quietly accepted.

What the review assesses

A launch review examines the plan as a document and the claims inside it. It looks at the brief, the proposed sites, the audience evidence, the lawful basis for any data used and the measurement method. It does not test panels, audit sites in person or verify footfall counts. Those are separate exercises.

Editorial judgement sits alongside evidence. A reviewer can say a plan is thin on measurement without having run the campaign, and should say so plainly. What a reviewer cannot do is present a vendor's own effectiveness figure as independent proof.

Nine checks before spend

1. The brief states the decision the campaign informs. If the plan cannot name what changes if the campaign works, the rest is decoration.

2. Each site claim has a source. Vendor pack, council record or third-party data. Name which one and its date.

3. The creative meets advertising rules. The CAP Bitesize guidance on ad rules applicable to OOH campaigns is the starting point for non-broadcast advertising, and the review should record which rules were checked.

4. Any audience data has a lawful basis. Targeting and measurement that touch personal data fall under the UK GDPR guidance and resources published by the ICO.

5. Effectiveness evidence is classed correctly. The data and marketing research published by the DMA is useful context, but it is industry research, not a verdict on your plan.

6. The measurement method matches the decision. Reach, context and response need different designs, as set out in our guide to out of home advertising service standards.

7. Ownership of post-campaign numbers is named. One person, one date, one format.

8. Contract exit terms are read. Notice periods and renewal clauses decide what a poor result actually costs.

9. The plan has a review date. Not a launch date. A date when someone re-reads the numbers.

For example, a team paying £400 a month for a small digital panel should expect the same nine checks as a national six-figure buy. The sums differ; the questions do not.

Separating claims from evidence

A launch review should label every substantive claim in the plan. Three labels work: verified, vendor-stated and unverified. Verified means an independent source is cited with a date. Vendor-stated means the supplier asserts it and the reviewer has not checked it. Unverified means nobody has sourced it at all.

Most plans arriving for review are heavy on the middle label. That is not automatically a problem, but the buyer should know which numbers they are taking on trust. The operations and delivery guide sets out how those labels carry through to delivery and reconciliation.

A short glossary helps when the review is shared with finance colleagues.

• Format: the physical or digital unit, such as a 48-sheet.
• Panel: one advertising unit in a defined location.
• Audience measurement: any method used to estimate who saw a campaign.
• Launch review: the pre-spend assessment of plan, claims and accountability.
• Post-campaign reconciliation: comparison of delivered activity against the plan.

Common questions

Does a launch review replace legal or data protection advice?

No. It records what was checked and by whom. Formal advice on contracts or data protection sits with qualified advisers, and the review should note where that advice was taken.

Can a reviewer sign off a plan they have not tested?

Yes, provided the limits are stated. A desk review can assess logic, sourcing and accountability. It cannot certify that a panel delivers what a vendor claims.

How long should a launch review take?

For a single-format local buy, half a day is realistic. Multi-format national plans with audience data need longer, because each claim needs a source and a label.

What happens to unverified claims?

They are flagged in the review and either sourced before launch or accepted in writing by the budget holder. Silence is not acceptance.