Credit risk gets a scorecard recalculated every night. Operational risk usually gets a spreadsheet updated once a quarter — despite sitting on more data than either. Including the one change that breaks the model.

A queue at a bank teller and a queue of aluminum brackets waiting on a welder are the same mathematical object wearing different overalls. Why flat two-week lead times are a guess, and utilisation math isn't.

Survival analysis tells you a customer is still alive. It says nothing about whether they're drifting toward Premium or Dormant. A transition matrix — built on the same RFM segments already on this site — fills in the gap.

Most branch closure lists are a spreadsheet ranked by last quarter's footfall. Survival analysis — the same hazard framework already running inside our churn model — asks a better question.

A REIT is a bank that collects rent instead of interest — and it has exactly the same blind spot. What the concentration-risk engine built for banking depositors reveals when pointed at a rent roll instead.

Netflix knows what you want to watch next. Why doesn't your bank know what you want to buy next? A look at recommendation engines, built from real RFM segmentation and uplift work.