What Uganda’s IVR Trial Shows About Practice-Based Scam Prevention

Digital financial fraud often succeeds in the moment between suspicion and action. A consumer may know that they should not share a PIN, trust an unexpected prize message or respond to pressure from someone claiming to represent a financial provider. The harder task is recognising the risk under pressure, pausing the interaction and choosing a safer next step.

An experiment led by Matthew D. Bird and Rafael Mazer in Uganda uses a randomised trial with 27,162 mobile money users to test whether an interactive phone-based audio fraud prevention game could improve real-world scam resilience. The intervention, Beera Mubangule (“Be Ready”), was developed in partnership with Viamo, IPA Uganda and the Gates Foundation funded Consumer Protection Research Initiative.

The Beera Mubangule results suggest that consumer protection programmes can be more effective when they move beyond warnings and build practical decision-making capability. Six to nine months after the intervention, trained users were 20% less likely to lose money to fraud, saving more through mobile money and filing more fraud complaints with the regulator, compared to the control group.

The case also fits into a wider policy discussion about fraud prevention in digital finance, where consumer education is increasingly being assessed alongside prevention, detection, disruption and recovery measures. Within that wider landscape, gamified and experiential approaches offer one way to move from general awareness towards practical behaviour change.

The Limits of Warnings Alone

Consumer-facing fraud prevention often relies on tips, alerts and reminders. These approaches assume that people can translate general advice into action at the exact moment a scammer applies pressure.

In digital finance, that assumption is often too weak. Fraudsters use urgency, impersonation and uncertainty to push consumers into fast decisions. A message that says “never share your PIN” may be familiar, but it does not necessarily prepare someone to handle a convincing call from a person claiming to represent a mobile money provider, telling them their account needs to be updated urgently, and they need their PIN for verification purposes.

The Uganda trial tested a different model. Instead of only telling users what to avoid, the intervention gave them a chance to practise recognising and responding to realistic scam scenarios before they faced them in real life.

A Fraud Prevention Game Designed for Basic Phones

Beera Mubangule was delivered as a short interactive voice story through Viamo and Airtel Uganda’s 161 information platform. Users accessed it for free on basic mobile phones in Ugandan languages, listening to realistic fraud scenarios and responding by pressing keys.

The story followed Sarah and Robert, a couple running a small business, as they encountered common scam situations. These included a request for a PIN, an impersonator claiming to be from a mobile money provider and a fraudulent prize offer. At key points in the user journey, users had to decide how the characters should respond and then heard the consequences of those choices. By the end of the game users are not just shadowing Sarah and Robert’s decisions, they are practicing their own.

The design rested on three practical principles:

Why Practice Matters

Key to the success of the intervention was that it required users to make decisions. That distinction matters because many scams succeed by narrowing the victim’s sense of time, choice and control.

Interactive storytelling creates a safer setting in which users can rehearse decisions before they face real consequences. It also allows feedback to be linked directly to the choice a person has made, making the lesson more concrete than a general warning.

Viamo’s explanation of gamification and behaviour change points to several features that help explain this effect: active participation, immediate feedback, real-world consequences, accessibility and scalability. These features are relevant to fraud prevention because they help convert passive awareness into a practical response routine.

The engagement data from Uganda is also notable. In the main fraud game, 93% of learners who started completed both levels, and 58% repeated the game more than once. This level of engagement is significant given that the game took an average of 18 minutes to complete.

Looking to contribute to collaborative anti-scam initiatives? GASA Working Groups bring members together to develop practical, real-world solutions.

What the Trial Found

The trial found that trained users lost 20% less money to fraud six to nine months after the intervention. They also saved more through mobile money and filed 67% more fraud complaints with the regulator.

Other behavioural outcomes help explain why the intervention matters beyond the headline loss reduction. The share of users reporting money lost to fraud in the previous six months fell from 7.1% to 5.7%. Among women, it fell from 8.3% to 5.7%.

The intervention also reduced risky practices. The share of people who reported sharing their phone, SIM or PIN information with others fell from 19.7% to 17.6%. The share of respondents checking whether unknown SIM cards were registered under their national ID rose from 21.2% to 24.9%.

The findings are especially relevant because the researchers also tested SMS reminders separately. The same fraud prevention messages did not produce meaningful protective behaviour among untrained users. Among trained users, however, reminders helped activate existing skills and translated into more complaint filing.

This distinction is important for anti-scam programmes. Reminders can be useful, but only when people already know what to do. Information may reinforce capability, but it does not automatically create it.

Consumer Protection Without Reducing Trust

Fraud awareness campaigns can sometimes create a concern that users will become more fearful of formal financial services. In digital finance, this risk matters because fraud can weaken trust in the same systems that are expanding access to payments, savings and credit.

The Uganda trial found the opposite effect. Trained users engaged more with mobile money, not less. They increased mobile money savings, used mobile money credit more and reported feeling safer on the platform.

For financial inclusion, this is an important lesson. Consumer protection does not have to reduce confidence in digital finance. When users can better distinguish legitimate contact from fraudulent contact, they may engage with greater confidence.

This is also why the intervention should be understood as more than an awareness campaign. It strengthened users’ ability to act safely within a digital financial system, rather than simply warning them that the system contains risks.

Why Delivery Costs Matter

The policy brief also highlights the role of delivery economics. Scaling the intervention to Uganda’s 11.4 million reachable mobile money users, at the observed completion rate, could prevent 160,000 to 205,000 fraud cases per year and save consumers USD 3.5 million to USD 4.5 million annually in direct losses.

The affordability of delivery was central to this level of cost-effectiveness. In the trial, low-cost access through existing mobile infrastructure enabled the IVR and SMS content to reach users without generating substantial delivery costs. At commercial telecom rates, however, the estimated annual cost of delivery increased from approximately USD 30,000 to USD 789,000, illustrating how delivery costs can become a major barrier to scale.

The Uganda experience highlights a broader policy lesson about the economics of scaling digital public-interest services. Sustainable delivery models matter as much as effective content. Governments, regulators, mobile operators and digital service providers should work together to reduce the cost of delivering evidence-based consumer protection services. This may include zero-rating where appropriate, but also other commercial or regulatory approaches such as sponsored access, preferential pricing or public-interest partnerships. Affordable delivery can determine whether proven interventions remain small pilots or become scalable consumer protection infrastructure.

Practical Lessons for Anti-Scam Stakeholders

The Uganda case offers several lessons for organisations designing scam prevention programmes.

Building Capability Before the Scam Happens

The trial shows that scam prevention can do more than warn people about risk. By placing users inside realistic scam scenarios, Beera Mubangule gave them a chance to practise the decision before facing it in real life.

That is one of the main values of the intervention. It did not ask users only to remember a warning. It asked them to pause, choose a response and see the consequence of that choice. The results suggest that this kind of practice can help users act differently later, including by losing less money, reporting more fraud and continuing to use mobile money with greater confidence.

For organisations designing consumer protection programmes, the case offers a practical shift in focus: from asking whether people received the message to asking whether they were better prepared to act when the scam arrived.

Sign up for the GASA newsletter to receive regular updates on scam prevention, research, and best practices.