Calling every retail advert retail media, and other planning mistakes you can observe

These retail media strategy mistakes are observable defects in a plan, not claims about how often businesses fail. Each can be checked before an England retailer exposes customers or advertiser data to a new placement.

Method: We selected defects that can alter a customer choice, evidence trail or go/no-go decision and that have a direct basis in the cited regulator, government or industry record.

Research date: 6 September 2026.

England and UK scope: The operating example is an England retailer. UK or Great Britain sources retain their stated geography; IAB material is an industry definition, not law.

Inclusions: On-site, off-site and in-store planning controls for an adult general-audience retail case.

Exclusions: Prevalence estimates, rankings, named suppliers, prices, campaign performance, regulated products and completed legal assessments.

Ranking: Non-ranked. The numbering is for reference only.

Conflicts: No commercial party funded the research or supplied private evidence.

1. Calling every retail advert retail media

A plan cannot assign inventory, data and measurement responsibility if its category is vague. IAB UK defines retail media around retailer or marketplace controlled advertising space, data assets and in-store opportunities. Correct the plan by naming the controlled property, advertiser exchange and on-site, off-site or in-store environment. Exclude ordinary merchandising and unrelated advertising.

2. Optimising before checking price and choice

A click goal can reward a presentation that conceals mandatory charges or steers an optional selection. The CMA's price transparency guidance covers total price, mandatory fees and taxes, drip pricing and partitioned pricing. Put product evidence, total-price presentation and freely made optional choices outside the optimisation score. A consumer-law failure stops the test.

3. Treating the ad label as a late creative detail

Recognition depends on the finished content and context. CAP Code section 2 says marketing communications must be obviously identifiable and addresses commercial intent and advertorials. Test the wording, location, size and timing in every format. A field marked "sponsored" in a specification does not prove that customers will notice or understand the rendered disclosure.

4. Equating usable tracking with permitted tracking

A technically complete exposure log does not establish the legal basis for a device operation or later profiling. The ICO's final storage and access technologies guidance applies PECR and, where relevant, UK GDPR to cookies, pixels, fingerprinting and similar techniques. Map each purpose and recipient, then test refusal and suppression. Privacy permission and analytical usefulness are separate verdicts.

5. Letting a commercial score cancel an access barrier

An inaccessible customer or staff task is not a small negative weight. The official disability guide for service providers explains the Great Britain duty to anticipate barriers and consider reasonable adjustments. Give accessibility its own reviewer, fixtures, evidence and stop authority. Include a supported alternative route rather than recording only automated checks.

6. Reporting movement as causal impact

Orders can change after a campaign for reasons unrelated to the placement. HM Treasury's current Magenta Book distinguishes observing an outcome from determining whether an intervention caused it. It is evaluation guidance, not a retail-media benchmark. Predeclare population, comparison, outcome window, missingness and stopping rule. Label results observed or attributed unless a qualified statistician accepts a credible counterfactual design.

Correct the plan before buying media

Assign an owner to each defect and preserve the failed evidence. Retest only the affected path with the same approved fixtures and versioned definitions. The decision remains HOLD until advertising, consumer, privacy, accessibility, security, IP, contract and measurement reviewers have issued separate verdicts for the actual proposition.