%202.png)
A side-by-side study of how digital fraud is reshaping trust, money and institutional confidence in South Africa and the United Kingdom, drawn from 1,335 real WhatsApp conversations. This report explores who loses money, who stays silent, who gets it back, and why the same fraud problem looks completely different depending on which economy it hits.
Methodology & data source
Based on 1,335 completed WhatsApp conversations, 705 in South Africa and 630 in the United Kingdom, 1,284 of them trust-screened and validated for analysis, this report runs the same instrument, the same way, in two markets that sit at very different points on inequality, unemployment, inflation and fraud regulation. That gap is what makes the comparison useful: where the two markets agree, the finding is probably human nature. Where they diverge, it is probably structural.
What you'll discover
The Long Con
The scams people resist are fast: 68% of South African near-misses and 78% of UK near-misses are caught in under 30 minutes. The scams that actually succeed look nothing like that, unfolding over days or weeks in 43% of South African loss cases and 63% of UK loss cases, while suspicion fades gradually instead of spiking.
The Fraud Burden Gap
South Africa's lowest-income fraud victims lost a median of 40% of one month's household income, against 7.5% for the country's highest earners, a 5.3× burden gap even though high earners lose money almost as often. Absolute loss and real harm are two different statistics.
The Inside-Job Suspicion Gap
64% of affected South Africans believe bank or institution staff are probably or definitely complicit in the fraud that hits them, against 17% of affected UK respondents, a belief that barely moves whether or not someone has personally lost money.
The Silence Gap
41% of affected UK respondents have never told anyone they were scammed, more than double South Africa's 17%. Combined with a lower reporting rate to institutions, a meaningful share of UK fraud never registers anywhere, not with a bank, not with a friend, not with a statistic.
The Confidence Gap
74% of South Africans and 88% of UK respondents already know scammers can use AI to clone voices and fake messages. Knowing barely changes doing: in South Africa, 10% of the AI-aware still react riskily to a too-good-to-be-true offer, against 12% of the unaware.
The Reporting Paradox
South Africans report fraud to their bank or another institution more often than UK respondents do, 47% versus 41%, despite believing far less that those institutions are acting in good faith. Reporting and trusting turn out to be two separate behaviours, not two names for the same thing.
The Barbell Effect
Loss incidence across South Africa's income bands does not fall in a straight line as income rises. It forms a barbell: the working poor lose money at 28%, the upper-middle class at 11 to 18%, and the country's highest earners at 27%, nearly matching the working poor.
Data highlights
About this research
Conducted by Yazi Research, a WhatsApp-native market research platform built on a simple observation: traditional research forces a choice between the depth of an interview and the scale of a survey. Structured questions, open text, voice-note storytelling and AI-moderated depth interviews run in a single WhatsApp conversation, so this study captures both quantitative scale and qualitative depth from the same respondents, in South Africa and the United Kingdom, at the same time.
Want the full report, or to run a similar rapid-response study with Yazi?
Book a demo →