Behavioral models that moved acquisition and engagement
At a major bank: a Markov model of app behavior, marketing-mix analysis, and PIX fraud detection — lifting activation and credit-lead quality.
At Banco Safra, growth depended on two questions: how to activate more app users, and how to attract the right credit customers — not just more of them.
More users, not better ones
Blunt acquisition and generic messaging brought volume, but activation lagged and credit-lead quality was uneven — while instant-payment fraud was a growing risk.
Model the behavior, then act on it
Renan built a Markov-chain model of app-user behavior to retarget communication to where it mattered, plus executive-level analyses that pivoted credit policy and marketing. He also led fraud-pattern analysis on PIX instant payments.
Activation and lead quality up
App activation rose +15% and high-quality credit-card leads rose +45% — with fraud patterns surfaced on the instant-payment stream.
What BIS would do
BIS builds behavioral and risk models with a decision attached — and reports the honest uncertainty around them. If acquisition, engagement, credit, or fraud is where your numbers live, this is the work.
Have a problem worth solving?
We start with a conversation, prove the value on a focused POC with your real data, then ship it — measurable ROI, not a model demo.