Littering is one of those persistent public problems that often feels unsolvable. Campaigns urging people to “do the right thing” rarely lead to lasting behaviour change, and enforcement is costly. But what if a behaviourally-informed digital intervention could shift public habits more effectively, and at scale? A recent study by Julia Kolodko, Karan Arora, Velina Hristova, Daniel Read, and Ivo Vlaev investigate with ‘Littergram’ – a digital platform where users photograph and report litter to their local councils. The results? Encouraging, and a clear signal that behaviourally informed digital nudges can nudge society in the right direction.
The problem: Engagement fatigue and the battle for attention
Littergram was created to empower citizens to take action against littering. Users snap photos of litter and tag the location, which then gets flagged to the local authority for clean-up. However, like many civic tech solutions, user engagement dropped over time. The team realised the challenge wasn’t just about awareness – it was about sustained motivation.
That’s where behavioural science came in. Instead of traditional awareness campaigns, the researchers deployed a six-week behavioural email intervention, incorporating research on psychological capabilities, social and environmental opportunities, and reflective and automatic motivations to inform the design of behaviour change strategies.
Using behavioural science to design an intervention
First, the research team adopted the Behaviour Change Wheel (BCW) methodology, a gold-standard framework developed by Michie and colleagues. At the heart of this model is the COM-B system – which stands for Capability, Opportunity, and Motivation as drivers of Behaviour.
The team began by identifying the target behaviour (posting pictures of litter on the Littergram app) and conducted a diagnostic survey to explore the underlying factors influencing that behaviour. This survey, based on the Theoretical Domains Framework (TDF), was completed by over 200 participants and provided rich insight into what enabled or hindered users from engaging with the app. The responses helped pinpoint three core behavioural drivers: behavioural regulation, emotions, and beliefs about consequences. These were then matched with suitable Behaviour Change Techniques (BCTs).
Our rigorous, step-by-step approach meant the digital intervention was not just intuitive – it was evidence-based and precisely tailored. Rather than rely on assumptions, the researchers used statistical modelling to identify which beliefs and psychological processes actually predicted posting behaviour. For instance, those who believed their actions had a meaningful impact on the environment or felt emotionally engaged were more likely to use the app. This allowed the team to design interventions that targeted these drivers directly. In doing so, the study not only reduced littering behaviour but also demonstrated how behavioural science, when applied systematically, can create scalable and effective digital tools for environmental change.
The three psychological domains identified – Behavioural Regulation, Emotions, and Beliefs about Consequences – reveal much about what drives environmental engagement. Behavioural Regulation points to the role of routines and self-monitoring in sustaining behaviour, suggesting that even simple reminders or cues can significantly increase participation. Emotions remind us that behaviour isn’t purely rational; users who felt emotionally connected to the cause were more likely to act. And Beliefs about Consequences highlight the importance of perceived impact — people are more likely to act when they believe their contribution genuinely makes a difference. These domains offered the behavioural “levers” used to craft targeted nudges in the intervention, translating complex theory into accessible, everyday action.
The intervention: Behaviour change in your inbox
We designed a targeted, three-part intervention, guided by the APEASE criteria (Acceptability, Practicability, Effectiveness, Affordability, Side-effects, and Equity) and mapped to a core behavioural domain: Social and environmental consequences (to shape beliefs), Monitoring emotional consequences (to tap into affect), and Self-monitoring of behaviour (to support regulation).
This minimalist approach balanced impact and cognitive load – avoiding message fatigue while maximising engagement. By anchoring each weekly message in a specific psychological lever, the campaign created structure, consistency, and clarity in its behavioural call to action. The emails sent to Littergram users were crafted with precision, with each new weekly message focused on specific Behaviour Change Techniques (BCTs) such as:
- Beliefs about consequences (“Litter harms wildlife and pollutes local environments”)
- Positive reinforcement (“Your efforts make your community cleaner”)
- Self-monitoring (“You posted 3 times this week – keep it up!”)
- Emotional engagement (“How did making a post make you feel?”)
Figure 1: Sample messages from the Littergram app
Source: Kolodko et al (2025)
The campaign didn’t guilt or shame users. Instead, it highlighted the impact of their actions, appealing to identity, progress, and social norms.
The results: A 62% surge in clean-up reports
The outcome? A 62% increase in daily posts to the app during the intervention (from 38.1 to 61.25 posts/day). While activity declined slightly after the emails stopped, post-intervention levels remained significantly higher than the baseline.
Figure 2: Engagement with Littergram before and after the intervention
Source: Kolodko et al (2025)
That’s no small feat. With no monetary incentives or app redesigns, the intervention managed to re-engage a large user base through strategic communication alone.
However, it wasn’t without challenges. The email open rate was around 25.7%, and unsubscribe rates hit 17%. Negative feedback-driven emails (“you didn’t post this week”) were more likely to lead to unsubscribes, reinforcing what behavioural scientists already know: too much pressure can backfire.
What we learned: Behavioural Science can scale
This project showed that theory-based digital interventions can work in the wild. It didn’t just improve metrics on an app; it hinted at how society can make small but scalable shifts in prosocial behaviour. Key lessons include:
- Digital nudging works – Especially when based on well-established models like BCW and built with empathy and insight.
- Motivation beats awareness – Users didn’t need to learn more about littering; they needed reminders that their actions matter.
- Tone matters – Positive reinforcement outperformed pressure-based prompts.
- Sustainability requires continuity – Once the emails stopped, behaviour dipped. Future campaigns might include spaced repetition, gamification, or social accountability to maintain impact.
Broader implications: Litter, health, and climate
While this intervention focused on littering, the implications stretch much further. The same behavioural scaffolding could apply to:
- Vaccination reminders
- Climate-friendly home upgrades
- Voluntary tax compliance
- Blood donation prompts
What’s powerful about this approach is its scalability. With minimal resources, a behaviourally informed email campaign can activate large numbers of people. Imagine if every civic tech platform, health app, or sustainability initiative embedded this kind of behavioural insight.
The takeaway: Nudge wisely
The UK’s battle with littering is far from over, but this study shows that with a bit of nudge-based wisdom, we can get a little closer to cleaner streets. The next frontier? Combining digital interventions with physical environment changes, community-based incentives, and real-time feedback loops.
Behaviour change isn’t just a theory anymore. It’s becoming one of our most powerful tools for tackling social and environmental challenges – one email at a time.
Find out more about the study here.
Ivo Vlaev is Professor of Behavioural Science at the Centre for Behavioral and Implementation Science Interventions (BISI), based at the Yong Loo Lin School of Medicine, National University of Singapore.
Daniel Read is Professor of Behavioural Science at Warwick University. Julia Kolodko has a PhD in Behavioural Economics from Warwick Business School, and is currently an adviser and trainer in applied behavioural science.