What creative has to do
Media buys the opportunity to be seen. Creative decides what happens next. In most categories creative is judged on one axis, whether it works. In gambling it is judged on two at once, whether it works and whether it stays inside rules that are stricter than almost any other sector's, and the second judgement is made by people outside the business: advertising standards bodies, regulators, platforms and media owners. Creative that performs and then gets banned has not performed.
This lesson treats the rules as a design constraint rather than a hurdle. The detail of the rules is in Marketing Compliance and Advertising Standards, and this lesson does not repeat it. The focus here is on the craft: the choice between offer-led and brand-led creative, building work that clears review the first time, testing creative honestly, and the evidence on safer gambling messaging in advertising, which is weaker and more interesting than most practitioners assume.
Offer creative and brand creative
Most gambling advertising falls into one of two families.
| Offer creative | Brand creative | |
|---|---|---|
| Message | A welcome bonus, free bet, price boost or promotion | Who the brand is, what the product feels like, why to choose it |
| Job | Convert people already in the market now | Make the brand the one people think of when they next want to bet or play |
| Typical channels | Paid search, paid social, affiliates, retargeting | Television, video, out-of-home, sponsorship, audio |
| Measurement | Fast and attributable | Slow, mostly measurable through tests and models |
| Regulatory exposure | High: significant conditions, accuracy, inducement rules | Moderate: appeal to minors, trivialisation, social or financial success |
| Main risk | Attracts bonus-motivated, low-value customers | Spends money that is hard to prove worked |
The wider advertising industry has studied the tension between these two for years. The IPA's report by Les Binet and Peter Field, The Long and the Short of It, explores, in the IPA's words, "the tension between long and short-term strategies for brands and businesses" and offers evidence-based recommendations on how to balance investment between them. Its central argument, that short-term activation and long-term brand-building do different jobs and that most advertisers need both, applies to gambling with one twist: the offer is the dominant form of activation in gambling, and offers are being regulated directly.
In Great Britain, the Gambling Commission capped bonus wagering requirements at ten times and banned promotions that require customers to take part in more than one type of gambling, with all changes in force from 19 January 2026. The industry code that the Commission tells licensees to follow already prevents pre-watershed television advertising from including sign-up offers targeted solely at new customers. Each restriction on offers shifts creative work towards brand and product, and operators that have relied on the size of the offer as their only differentiator find they have nothing else to say.
Building creative that clears
Creative that is rejected late costs twice: the production budget and the media slot it was meant for. The cheapest compliance is the kind designed into the brief.
The most consequential rule for creative in Great Britain is the "strong appeal" test that has applied since 1 October 2022. The ASA's guidance states that gambling ads "must not be of strong appeal to children or young persons", and that top-flight footballers with considerable under-18 followings on social media are likely to be high risk. The test operates on the material, not on the intention or the media plan, so it has to be applied at the casting and storyboard stage, not at the edit.
Prominence is the second recurring failure. In a 2024 ruling against a social casino advertiser on TikTok, the ASA upheld a complaint where the disclaimer that "In-game rewards are not real money" appeared in small white text at the bottom of the advertisement. A mandatory or qualifying statement that the viewer cannot read in the time available is treated as absent. Short-form vertical video, which is designed to be watched with the thumb already moving, is the format where this goes wrong most often.
The third failure is forgetting that brand content is advertising too. Gambling brands' social accounts post memes, opinions on the weekend's matches and editorial-style commentary, and the teams writing them often do not think of themselves as producing advertisements. The ASA has made clear that if the purpose of a post is "to 'sell something'" and the content is "directly connected" to the product, it is regulated under the code, which brings in the strong appeal test and the restriction on people under 25 appearing in gambling advertising on social media. In practice, an always-on social team produces more gambling advertising by volume than the campaign team, with less review. The brief and the review tiers below should apply to its output as well, scaled to the risk, because a meme built on a youth-culture reference can fail the strong appeal test as surely as a television commercial.
A creative brief with the constraint built in
A brief that gets creative through review the first time contains, in addition to the usual objective, audience and proposition:
| Brief section | What it specifies |
|---|---|
| Markets and media | Every market and placement the asset will run in, because the rules differ |
| Talent and characters | Age and audience profile of any person or character, checked against strong appeal tests before casting |
| Visual references | Excluding styles associated with youth appeal, such as animation styles and gaming aesthetics, unless reviewed |
| Claims | Every factual claim with its substantiation, held before production |
| Offer | Exact current terms, significant conditions to appear with the claim, and expiry date of the asset |
| Mandatory elements | Age and safer gambling messaging with minimum size and duration per format |
| Formats | Each cut-down listed, because a compliant 30-second version can produce a non-compliant 6-second one |
| Review tier | Who signs off and when, so review is scheduled rather than discovered |
Notice what is absent: a section headed "compliance" at the end. The constraint is distributed through the brief because that is where the decisions are made.
Testing creative honestly
Most creative testing in gambling is done inside the ad platforms: two or more variants run against the same audience, and the platform reports which won. It is fast and useful, and it is easy to get wrong in three ways.
Optimising to the wrong event. A variant that maximises registrations may attract people who never deposit, and one that maximises first deposits may attract people chasing the offer who never return. Tests should be read on a value measure, at minimum first deposit plus net revenue at 30 or 90 days, recorded in the operator's own data.
Letting the platform pick the audience. Many platform tests let the delivery system send each variant to the people most likely to respond to it, which means the variants are seen by different people and the comparison is not clean. Where the platform offers a randomised split, use it.
Reading noise as a result. Small differences in conversion rates on modest budgets are usually noise. Lesson six explains why advertising effects are hard to measure precisely; creative tests are the same problem at smaller scale. A practical rule is to agree, before the test starts, the minimum difference worth acting on and the sample needed to detect it, and to stop tests on that basis rather than when one variant happens to be ahead.
A worked test
The figures are hypothetical. An operator tests two paid social videos with equal spend of £50,000 each, randomised across the same audience.
| Variant A: offer-led | Variant B: product-led | |
|---|---|---|
| Registrations | 2,000 | 1,400 |
| First-time depositors | 900 | 750 |
| Cost per first-time depositor | £56 | £67 |
| Net revenue per depositor at 90 days | £70 | £110 |
| Net revenue at 90 days | £63,000 | £82,500 |
| Bonus cost (£30 per depositor for A, none for B) | £27,000 | £0 |
| Net revenue less bonus cost at 90 days | £36,000 | £82,500 |
On registrations or cost per depositor, A wins clearly and would be scaled. On value net of the offer, B is worth more than twice as much. The test changed nothing about the creative; it changed what the team chose to count.
Two further checks belong in the readout. First, is B's higher value spread across many customers or concentrated in a few heavy players? If a handful of accounts account for most of the difference, the "winning" creative may be finding people who gamble heavily, and lesson six explains why that is a risk rather than a result. Second, run the variants through the same compliance review at scale that they received at production. Winning creative gets more spend, more formats and more cut-downs, and each is a new opportunity for a prominence failure.
Safer gambling messaging in advertising
Most gambling advertising now carries safer gambling messaging, and a growing share of advertising space is given over to it entirely. In Great Britain, the industry code requires most operators to direct 20% of all eligible advertising, broadcast and digital, to safer gambling messaging, a requirement extended from broadcast to digital from 1 December 2023. Entain's 2025 annual report, for example, states that it continued to spend 20% of its UK advertising budget on safer betting and gaming communications.
The evidence on whether the messages work is more cautious than the volume suggests.
The best-known tagline did not change behaviour in experiments. A study published in The Lancet Public Health in 2022 by Newall and colleagues tested the long-running British message "When the fun stops, stop" on online bettors and found that the message did not significantly reduce betting compared with a control group. The industry has since replaced it with the "Take Time To Think" campaign.
Better messages can be designed and tested. A 2024 study for GambleAware by YouGov, the Behavioural Insights Team and the University of Bristol tested alternative safer gambling messages on operator advertising and found that "several messages performed better than 'Take Time To Think'", with clearer signposting to advice and support increasing the likelihood of people searching for help.
Some safer gambling advertising backfires. A 2025 randomised controlled trial for GambleAware tested operator safer gambling videos and found that some had a backfire effect, with presenting gambling as "harmless fun" a likely driver, and that young people and those already experiencing harm were most susceptible.
The Responsible Gambling course covers the wider evidence on harm prevention. The practical conclusions for a creative team are specific.
- Treat safer gambling creative as creative. It needs a brief, an objective (use of a deposit limit, a time-out, a visit to support), and a test, not a logo and a line added at the end.
- Do not let it become brand advertising in disguise. A safer gambling advertisement that mostly shows people enjoying the product is the format the 2025 trial found could backfire.
- Signpost something concrete. Messages that point to a specific tool or source of help performed better than generic exhortation.
- Measure behaviour, not recall. The question is whether customers exposed to the message set limits or seek help more often, which an operator can test in its own data with a holdout.
- Keep the mandatory elements legible. Whatever the tagline, the age and safer gambling messaging required by the code must pass the prominence test in every format.
A creative system rather than campaigns
Operators that do creative well in this sector tend to run a system rather than a series of campaigns. A versioned library of approved claims, offers and mandatory elements per market lets teams assemble compliant work quickly. Every asset carries an expiry date tied to the offer terms it describes, so that nothing runs after the terms change. Creative performance is read on value, net of offer cost, from first-party data. Wear-out is monitored, because gambling audiences see a great deal of advertising and response falls as the same execution repeats. And the library is updated whenever a ruling, a platform change or a regulatory change alters what can be said, so that the next brief starts from the current rules rather than last year's.
That system is also what makes the next lesson possible. Measuring media properly depends on creative being consistent enough, and documented well enough, that a test of a channel is not secretly a test of whichever execution happened to run in it.