How UK agencies check paid-social adverts before launch

What to take away

What the CAP Code covers before a paid-social ad is released

The CAP Code is the UK advertising rulebook written by the Committee of Advertising Practice and enforced by the Advertising Standards Authority. It applies to paid social ads shown to British audiences, whether the advertiser sits in London, Manchester or overseas. The rules cover claims, targeting, disclosures, recognisability, pricing, comparisons and harm.

They sit alongside the advertising codes, which include the broadcast code and the CAP Code itself. A campaign that runs on Meta and ITV must satisfy both sets.

Platform policy is not the same as the CAP Code. Meta Ads Manager may approve an ad that the ASA later rules against.

The CAP Code applies in England, Scotland, Wales and Northern Ireland. A social media advertising agency should check the audience and any additional brief requirements before release; see the related Welsh public-sector guide.

A campaign aimed at Greater London, the North West or Yorkshire still sits under the same CAP Code. Record the intended audience and any local brief requirements in the pre-release check.

Other regulators overlap. The Information Commissioner's Office enforces UK GDPR on data used for targeting. The Competition and Markets Authority watches online advertising and dark patterns.

Substantiation of claims in social media advertising

Every objective claim in a paid social ad needs evidence before release, including price, performance, health and comparative claims. The CAP Code requires that advertisers hold documentary evidence before publication, not after a complaint.

A claim like "clinically proven" needs a trial that supports the exact wording, population and outcome. A claim like "cheapest" needs a market-wide price check.

Substantiation is a pre-release task, not a defence. The evidence must be in the advertiser's possession when the ad goes live. If the ASA asks, the advertiser must supply it quickly. A social media advertising agency should keep a substantiation file for each campaign, with the source, date and the sentence it supports.

The quality checklist sets out how claims, evidence and data access fit together: claims, access, data.

Social proof is a common trap. Reviews, ratings and user counts are claims. "Rated 5 stars by 10,000 customers" needs a verifiable source and a date. Influencer posts that include a claim bind the brand as well as the creator.

A creator saying "this cured my eczema" is a health claim, and the brand can be held responsible if it controlled the brief.

Comparative advertising needs care. A direct comparison must compare like with like, be verifiable, and not mislead by omission. "Better than the leading brand" needs a named basis and evidence. Price comparisons must include the same specification and date. A claim that a product is "free" must be genuinely free, with no hidden cost.

Financial and health claims attract extra scrutiny. A crypto ad cannot promise returns. A supplement cannot claim to treat disease without the relevant authorisation. A weight-loss ad cannot show unrealistic results.

Some product categories have additional advertising-code restrictions, so check the rules that apply to the campaign. See Advertising codes - ASA | CAP.

Targeting rules and vulnerable audience protections

The CAP Code does not set every targeting rule, but it does require that ads are not directed at people who cannot be expected to understand or resist them. Children and young people need protection.

Ads for age-restricted products must not be targeted at under-18s, and platforms' own age gates are not enough on their own. A paid-social campaign for alcohol, gambling or e-cigarettes should exclude under-18 audiences explicitly, not rely on a default.

Vulnerable adults also need protection. People with mental health conditions, financial difficulties or cognitive impairments may be less able to assess an offer. The CAP Code requires that ads do not exploit their vulnerability.

A debt ad aimed at people with poor credit, or a gambling ad aimed at high-risk players, can breach the rules even if the targeting was technically legal.

The ICO's UK GDPR rules on data used for targeting sit alongside this.

Targeting based on sensitive data is restricted. Health status, sexual orientation, religion and political views are special category data under UK GDPR. Using them for paid-social targeting without a lawful basis is a data breach and can also breach the CAP Code.

Lookalike audiences built from sensitive data carry the same risk. The CMA's online advertising and dark patterns guidance adds further limits on manipulative targeting.

Before launch, record the geographic audience and make sure the ad's language and offer match the brief.

For regulated products, check any product-specific requirements for the target audience before release. Do not assume one audience setting covers every campaign requirement.

Recognisability, ad labelling and disclosure of commercial intent

Paid social ads must be obviously identifiable as ads. The CAP Code requires that marketing communications are clearly distinguishable from editorial content. On social platforms, the ad label must be visible, the brand must be identifiable, and any commercial relationship must be disclosed.

A TikTok video that looks like a creator's organic post still needs a clear ad label and a disclosure in the caption.

Influencer marketing is a common failure point. Put the disclosure upfront and make it unmistakable; do not rely on a brand tag at the end or bury "#ad" among other hashtags.

"Paid partnership" tools help, but they do not replace a clear label in the content itself. A social media advertising agency should brief creators on the exact wording and placement.

Ad disclosure recognisability is not just about a label. The overall impression matters. An ad that mimics a news article, a review site or a user's personal post can mislead even with a small label. Native advertising must not disguise its commercial intent.

A fake countdown timer, a fake review or a fake "sold out" message can breach the CAP Code and the CMA's dark patterns guidance.

Disclosure also covers who is behind the ad. A political ad, a financial promotion or a health claim may need the advertiser's legal identity.

The ASA can act on a complaint about a missing disclosure, whether the ad ran on Meta, TikTok or Google.

Using ASA rulings in a pre-release review

Published ASA rulings can help an agency see how the CAP Code has been applied to real ads. Review decisions relevant to the campaign's claims, targeting and presentation as part of the pre-release check.

The ASA upheld complaints about Ryanair's "Jab and Go" ad, ruling that the airline's claim was misleading, and ordered that the ad must not appear again.

Use the ASA's published rulings search to find decisions relevant to the campaign's sector and issues. Read the full decision before translating it into a pre-release check; a ruling on one ad is not a substitute for checking the draft at hand. The full list is searchable: Rulings - ASA | CAP.

Agencies should read rulings in the sectors they work in. A ruling against a competitor is a free compliance lesson. The ASA publishes the ad, the complaint, the advertiser's response and the decision. That structure shows what evidence the advertiser failed to provide. It also shows how the ASA interpreted the CAP Code in that context.

The ASA can rule on an ad that is no longer running. That means a campaign cannot be made compliant by deleting it after a complaint.

Repeat rulings escalate. An advertiser that keeps breaching the CAP Code can be referred to the ASA's sanctions process. That can lead to naming on the non-compliant online advertisers list. A social media advertising agency should track rulings in its client's sector and feed them into the pre-release check.

Copy Advice and the pre-release clearance route

The ASA Copy Advice service is a confidential pre-release check. It gives an opinion on whether a draft ad is likely to breach the CAP Code. It is not a binding clearance, but it is a useful risk check. Agencies can use it for a new claim or format.

The service is separate from the complaints process and does not stop a later complaint. It is explained here: Copy Advice - ASA | CAP.

Copy Advice is not a substitute for legal review. It does not cover UK GDPR, consumer law or sector-specific rules. A financial promotion still needs the relevant authorisation. A health claim still needs the relevant evidence. A gambling ad still needs the relevant licence. Copy Advice sits alongside those checks, not above them.

Use Copy Advice early. A draft already in production is harder to change, and a claim central to the campaign may need rewording or dropping. The earlier the check, the more options the agency has. A social media advertising agency should build Copy Advice into the creative brief, not bolt it on at the end.

For complex campaigns, combine Copy Advice with a legal read. A campaign that touches health, finance, children or vulnerable adults should get both. The pre-release check should record what was asked, what was advised and what was changed. That record is useful if a complaint arrives later. It also helps the next campaign.

Sanctions and the non-compliant online advertisers list

When the ASA upholds a complaint, it publishes the ruling and asks the advertiser to change or withdraw the ad. Most advertisers comply. Those that do not can face sanctions. The ASA can ask platforms to remove paid ads, ask search engines to remove links, and name the advertiser publicly.

The sanctions process is set out here: Sanctions - ASA | CAP.

The non-compliant online advertisers list names advertisers that have not complied with ASA rulings. The list is public and can affect brand reputation, platform access and retailer relationships. It is maintained here: Non-compliant online advertisers - ASA | CAP. Being named is a serious outcome for a brand, and it can follow the agency too.

Sanctions can escalate. The ASA can refer persistent offenders to other regulators, including trading standards and the CMA. It can also work with platforms to remove ads at scale. For a social media advertising agency, the risk is not just the client's reputation. It is the agency's own record and its relationship with platforms and clients.

A pre-release check is a practical way to reduce the risk of a ruling, removal or public naming. The agency should treat the sanctions list as a live risk register. If a client's sector appears on the list, the pre-release check should be tighter.

Building a pre-release check into agency workflow

A pre-release check should be a gate, not a formality. It should sit between creative approval and campaign launch. The check should cover claims, targeting, disclosure and evidence. It should have a named owner and a recorded outcome. The agency should be able to show the check to a client, a platform or the ASA.

A simple workflow helps. The steps below fit most paid-social campaigns and can be scaled for larger ones.

Use the same seven steps for every campaign, not just the risky ones.

A worked example shows how this runs. A drinks brand plans a paid-social campaign for a new low-alcohol product in England and Scotland. The claim "less than 0.5% ABV" needs a lab report. The targeting excludes under-18s. The ad label is clear.

The influencer brief specifies "#ad" at the start of the caption. Copy Advice is used because the claim is new. The evidence file holds the lab report, the targeting screenshot and the Copy Advice reply. The campaign launches with a recorded check.

The workflow must cover the whole UK, not just England. Record the countries and audiences included, and check the applicable campaign requirements before release.

A social media advertising agency should check any regulatory layers relevant to the campaign, rather than assume one review covers every issue. See UK paid social rules.

A final read of the full pre-release process helps: Checking a paid social advert.

Common questions

Where the basis is not obvious, the ad should state it, for example the category, the market and the period. A claim of "best-selling" across the UK needs sales data for that market, not one retailer's figures.