Before an advert leaves the creative tool, the checks that decide whether it runs

An approved master has been sent to a creative tool for a UK non-broadcast display advert. The builder can resize it, generate text and combine assets into different versions. Before release, a reviewer must inspect those exports as the audience will see them, rather than treating the source file or a software verdict as approval.

1. The display advert enters the export queue

For this worked scenario, the team has one product advert with a claim and an offer. Its intended medium, placement, audience, product and UK geography are written down first. That information identifies the rules to check and prevents a verdict for one channel being copied onto another.

The reviewer confirms that the material falls within the CAP Code's published scope. The CAP Code covers non-broadcast advertising, is written by the Committee of Advertising Practice and is enforced by the Advertising Standards Authority. Platform, contractual or sector-specific requirements still need their own owner.

2. The reviewer examines what each placement shows

The reviewer opens the actual exported versions, including the first frame, crop and collapsed view. This matters because automated resizing or a different asset combination can alter prominence and implication, even if the master looked clear when approved.

In this scenario, the check finds that the commercial label is present in the master but does not remain clear in every rendered view. The offer callout also becomes less legible in a smaller placement. Those versions are held for correction; the source file cannot rescue a label hidden after more or lost in a crop.

3. The claim is matched to evidence

Next, the reviewer underlines factual statements, comparisons, demonstrations, superlatives, testimonials and implied results. For each one, the record names the evidence owner, date, relevant population and limitations. The visual is checked too, because it may imply more than the words say.

Here, a generated suggestion has supplied a line that sounds like an objective claim, but no supporting material is available for review. The release decision is therefore to remove or narrow that wording until evidence can be produced. Under CAP Code section 3, objective claims need documentary support before publication; generation is not substantiation.

4. The offer is read from the buyer's view

The reviewer checks the product, quantity and eligibility, then looks at total price, mandatory charges, expiry, availability and significant restrictions at the size actually served. Any comparison must have an identifiable and appropriate basis. The wording and presentation are considered together, not as isolated fields in the builder.

The scenario's offer is not released until those points are clear in the export. CAP Code rules 3.17 to 3.22 and the CMA's Pricing Practices Guide cover price statements, drip pricing, reference prices and time-limited offers. The team also captures the rendered version for each materially different placement, since a correct master can become unclear when its callout shrinks or moves.

5. The rule route is widened where needed

Before sign-off, the reviewer checks whether the audience or subject brings additional rules into play. Children, vulnerability, health, finance, gambling, alcohol, environmental claims and safety demonstrations need attention beyond the general code. The team uses the relevant sector regime instead of assuming that a general review settles every issue.

For this scenario, the product category and audience are checked rather than presumed.

Depending on what they show, the relevant checks can include children's rules and platform policies; the Portman Group's alcohol code; Gambling Commission licence conditions and codes of practice; the Human Medicines Regulations 2012 and MHRA Blue Guide; FCA approval under section 21 of the Financial Services and Markets Act 2000; or the Green Claims Code.

The advertising guidance library helps locate material but does not pre-approve an advert.

If the piece were for television, radio or on-demand, the team would route it to the BCAP Code rather than reusing the non-broadcast verdict. Ofcom is the statutory backstop; Clearcast clears television and video-on-demand copy, while Radiocentre clears radio copy before transmission. A landing page or influencer video can need different recognition, space and presentation decisions despite sharing a message.

6. The reviewer records a decision and a correction route

The reviewer saves the final copy, imagery, evidence pack, format preview and response under one version number, then marks the result approved, revise or stop. For this advert, the unclear exports and unsupported line mean revise. A qualified person makes that decision; the tool does not issue a compliance certificate.

Before another version is released, the team tests the least forgiving combinations: the longest approved headline, smallest placement, weakest colour contrast, highest price and most restrictive condition. It checks that an endorsement cannot be paired with the wrong product or a substantiation note with a different claim. The operational controls must also let an owner stop delivery, replace the asset and preserve the served version if evidence expires or an error appears.

A claim, offer, crop, audience rule, destination or evidence change reopens review, so an old approval cannot silently transfer to a new variant. If precedent informs the reasoning, a dated ASA ruling remains a fact-specific illustration, not a substitute for checking the present code. Non-broadcast advertising is not pre-cleared; CAP Copy Advice can review draft copy in confidence before release.

Rules carried into the next build

The scenario leaves a short set of release rules: review the export, match each claim to evidence, inspect the offer in its served size, confirm the relevant code route and record who can correct or withdraw the work.

These checks keep an automated production path reviewable without asking software to make the final judgement.

When a new export changes the message or presentation, run the review again rather than carrying forward the earlier decision. The disclosure policy should make recognition expectations clear, and the CAP rules reference can support the code check. A reliable correction route matters as much as a fast creation workflow.