The gambling industry’s messages such as “gamble responsibly” frame gambling harms as a matter of individual responsibility. Counter-industry warnings such as “gambling products are designed to be addictive” instead highlight systemic drivers of harmful gambling. In this blogpost, Maira Andrade and Philip Newall draw on their mixed methods research with 4,000 UK gamblers to show how policymakers could best use such counter-industry warnings.
The problem with “Gamble Responsibly”
“Gamble responsibly” is a warning familiar to many people from gambling advertising. But the gambling industry’s focus on individual-level solutions reflects a recurring pattern in behavioural science, by neglecting potentially more effective system-level changes, such as banning gambling advertising.
Evidence from Britain, which has had a deregulated gambling market with widespread gambling advertising for nearly two decades is clear. Harms are not confined to the 1–2% of the population usually classified as “problem” gamblers (in itself a stigmatising individual-level frame). People who suffer from subclinical harms (meaning, below the “problem” threshold) can experience up to half of all gambling-related harms. In Great Britain, low- to moderate-risk gamblers accounted for 11.9% of surveyed adults in 2024. UK policy has started to reflect this evidence through measures such as £5 and £2 stake limits, and restrictions on harmful product design features.
This is why industry-led harm prevention messages are no longer enough. Slogans such as “gamble responsibly” and “when the fun stops, stop” emphasise personal responsibility, can lead to further stigma, and are often ignored by gamblers. They have also been shown to have no positive impact on gambling behaviour. In 2023 Australia became the first country to replace industry slogans with independently-designed messages, followed by Britain where the transition has been announced for 2026. Given the inherent flaws of the i-frame model for harm prevention messages, other countries like the US and Canada must catch-up soon.
Flipping the script: Counter-industry messages
Counter-industry messages are designed to highlight harmful industry practices, and have been successfully implemented across several public-health domains. The anti-tobacco ‘Truth’ campaign aimed to expose the tobacco industry’s deceptive advertising strategies, and was linked to a 22% decrease in youth smoking in the US between 1999 and 2002. In the food industry, messages highlighting how junk food marketing is incompatible with adolescent values were shown to significantly reduce positive associations with unhealthy food and better dietary choices. Finally, counter-industry messages bringing attention to the harms of alcohol sports sponsorships were found to significantly increase harm awareness, and intentions to reduce alcohol consumption.
In the UK, the “Odds are: they win” campaign was launched by the Greater Manchester Combined Authority with counter-industry messages such as “The main purpose of gambling companies is to maximise profit, generated through customer losses”. Although a later report showed the campaign achieved excellent online engagement, effectiveness was not quantitatively assessed. Therefore, interventions that quantitatively assess counter-industry gambling campaigns remain unexplored, creating a significant evidence gap.
Design and evaluation of counter-industry messages
We set out to address this gap in a recent study (co-authored with Archie Spicer, Leonardo Weiss-Cohen, Simon T van Baal, Jamie Torrance, and Leon Xiao). We used quantitative and qualitative methods to evaluate 10 counter-industry messages in a sample of 4000 UK-based gamblers. Half of the messages were selected from previous campaigns and research, including Gambling Understood, Chapter One, ‘Odd are: They win’, and Torrance et al. (2025). The remaining five messages were developed in collaboration with gambling researchers and lived experience advocates. For each message, participants used a Likert scale to rate how relevant the message was to them, how relevant it was to those experiencing gambling harms, if it made them want to gamble less, and if they believed it.
Which messages performed best?
On average, participants believed all messages, which is an important finding given the uniqueness of framing gambling harms as a matter of corporate irresponsibility. Seven messages made participants want to gamble less, and five were rated as personally relevant.
Using the Likert scores, we ranked the messages by overall performance (see Figure 1). The three most effective messages were already part of existing harm-prevention campaigns. The highest-rated message overall was “The main purpose of gambling companies is to maximise profit, generated through customer losses”. However, a natural-language-processing assisted qualitative analysis showed that “Gambling products are designed to be addictive”, despite ranking third, received the most positive responses for being direct, short and clear, while the best-scored message received criticism for being too long. The free-text analysis also showed that personal responsibility views were highly prevalent among the participants, reinforcing the potential benefits of counter-industry narratives in harm prevention messages.
Figure 1: The 10 counter-industry messages and their abbreviations as ranked by overall performance.
Source: Spicer, Andrade et al (2026)
Who found the messages most personally relevant?
To assess how messages perform across different risk profiles, we used the Problem Gambling Severity Index (PGSI) to measure harmful gambling, where answers are scored between 0 and 27. Overall, 14.6% of participants (n = 598) were categorised as high-risk gamblers (i.e., score of 8 or above). Participants rated all 10 messages as more personally relevant as PGSI scores increased, indicating that the counter-industry messages resonated most with those experiencing greater harm.
A message for every audience
Harm prevention messages can utilise different behavioural change techniques. The seven messages designed by the Australian government, for example, include self-appraisal (“Think. Is this a bet you really want to place?”), loss-based (“Chances are you’re about to lose”), and positive-emotional messages (“Imagine what you could be buying instead”).
Our counter-industry study is one in a series of investigations testing the effectiveness of different message types. To assess how different messages may appeal to different risk profiles, we combined data from all our experiments to compare counter-industry, loss-based (e.g., “99% of Gamblers Lose in the Long Run”), self-appraisal (e.g., “Would you be comfortable telling your family how much you lost gambling today?”), and positive emotional messages (e.g., “Quitting gambling can help you with the relationships that matter the most to you”).
How message evaluation can inform optimal implementation
Figure 2 shows the results of the combined analysis, and suggests that counter-industry and loss-based messages may be ideal for population-level campaigns, such as public transport adverts, while self-appraisal and positive emotional messages could be best suited for high-risk environments, such as in-game pop-up notifications. This is because population-wide campaigns mostly reach lower risk gamblers, while in-game pop-ups are more likely to reach individuals experiencing severe harm.
All messages felt more personally relevant as a participant’s risk of gambling harm increased (i.e., higher PGSI scores). However, counter-industry and loss-based messages began to feel personally relevant even to gamblers categorised as low-to-moderate risk (PGSI <8). In contrast, self-appraisal and positive emotional messages resonated more with individuals already experiencing greater gambling harm.
Figure 2. Results of a combined analysis comparing effects of PGSI interactions between counter-industry, loss-based, positive emotional, and self-appraisal messages.
Source: Newall et al. (2026)
Looking ahead: What this means for policy
The UK’s 2026 roll-out of independently-designed messages puts an end to decades of industry-controlled narratives framing potential gambling harm as a matter of personal responsibility. Here we demonstrate how counter-industry messages can shift harm prevention messages from an i-frame of individual responsibility toward an s-frame of systemic accountability. Importantly, our overall findings highlight how policymakers may consider multiple communication strategies when designing harm prevention messages, and the importance of implementing both quantitative and qualitative methods of evaluation across different risk profiles. Abandoning industry narratives in favour of evidence-based messages can be a significant step for harm prevention in the UK and across other jurisdictions that follow suit.
Click here for the full article in Behavioural Public Policy.
Maira Andrade is a PhD student at the School of Psychological Science at the University of Bristol. Her research focus is on gamblification and the intersection of online gambling and new technologies, in particular cryptocurrencies.
Philip Newall is a Senior Lecturer at the School of Psychological Science at the University of Bristol. Their research is well known for applying concepts from behavioural science to issues of gambling psychology and policy.