01
Behavioral research with affected customers
30 in-depth interviews, 60 to 75 minutes each, with customers whose accounts were closed over suspected money mule activity.
False positives in fraud prevention: customer context in risk decisions
A digital bank wanted to understand what lay behind accounts closed over suspected money mule activity, going beyond the transaction signal. Sensorama ran behavioral research with 30 affected customers and cross-checked their stories against the bank's internal data, to support product and fraud prevention decisions.
- Fraud Prevention
- Anti-Fraud
- Behavioral Research
- Qualitative Research
- UX Research
Challenge
The fraud prevention model relied mostly on transaction signals, with little visibility into customer context. When legitimate behavior is read as fraud, the loss is threefold: revenue, customer trust, and operational, legal and reputational costs. The bank needed evidence to reduce false positives without losing its ability to stop real fraud.
30
INTERVIEWS
in depth with affected customers
3
MONTHS
of project work, March to May 2026
3
FRONTS
of guidelines for fraud prevention
11
RECOMMENDATIONS
organized across the three fronts
What we delivered
02
Cross-check with internal data
Customer stories were compared with the bank's behavioral and transactional data, to reduce bias and place what the system read as risk next to each customer's context.
03
Risk segmentation matrix
Criteria to classify each case from the interview and the customer's history at the bank, for a more accurate reading of intent and risk exposure.
04
Guidelines on three fronts
Risk models more sensitive to context, progressive interventions before irreversible actions, and preventive education at moments of risk.
05
Prioritized recommendations
11 recommendations across the three fronts, organized by prevention stage (prevent, detect, analyze and investigate) and by priority.
We listened to the people affected by the risk decision
Sensorama ran 30 in-depth interviews between March and May 2026 with customers whose accounts were closed over suspected money mule activity. Sensorama led the research as a third party, without naming the bank during recruitment, to reduce defensive answers.
Customers whose accounts were closed over suspected fraud
Adults aged 18 to 55
Accounts closed between December 2025 and April 2026
Customers in Brazil
How the research was run
Behavioral research validated against the bank's systems, to build evidence and reduce bias.
One-on-one interviews with projective techniques
Unbranded research, led by Sensorama as a third party
Stories cross-checked with behavioral and transactional data
Anonymization and legal and compliance review
By the end of the project, the bank had an evidence base for product decisions and guidelines on three fronts to improve fraud prevention, with one shared principle: understand better before acting.
Want to reduce false positives without loosening your fraud prevention?
We run behavioral research that brings customer context into risk and product decisions.
Let's talk