Risk-based pricing programme worth £20M/year unlocked
Customer gaming patterns quantified for the first time
Time to insight reduced from weeks to minutes
Contract re-entry terms tightened based on survival evidence
An interactive survival analysis tool gave a UK insurer its first granular, explorable view of customer lifecycles, directly enabling a risk-based pricing programme worth £20M per year. The tool exposed previously invisible gaming patterns, where policyholders cycled in and out of policies to exploit introductory discounts, and provided the shared evidence base that accelerated contract change decisions from quarters to weeks.
The Problem
A large UK insurance business had no shared, quantified view of how its customers behaved over time. Retention was discussed in broad strokes at board level, but nobody could answer precise questions: which acquisition channels produced customers who stayed longest? At what points in the contract cycle did lapse rates spike? Were certain product structures incentivising customers to leave and rejoin on cheaper terms?
The Solution
We built an interactive survival analysis tool that gave the business its first granular, explorable view of customer lifecycles across every meaningful dimension. The core was a set of Kaplan-Meier survival curves computed across the full policy book, segmented by acquisition channel, product type, contract terms, renewal history, and demographic cohort.
Policy administration data was joined with claims history, renewal records, and channel metadata to create a single longitudinal view of each customer's journey. Every policy start, renewal, lapse, cancellation, and rejoin event was timestamped and linked. Careful attention was paid to defining "survival" correctly: a customer who lapses and rejoins three months later on a new introductory deal is not the same as one who renews continuously. Capturing re-entry events explicitly turned out to be essential for identifying gaming.
Rather than delivering a static report, we wrapped the analysis in an interactive dashboard with cohort selection, curve overlays, adjustable time horizons, and a scenario modelling module. Pricing and product teams could ask "what if" questions: what happens to the survival curve if we remove the introductory discount for returning customers? What is the revenue impact of extending the minimum contract period? Each scenario adjusted the survival estimates based on behavioural assumptions calibrated from historical data, giving decision-makers a quantitative basis for contract changes rather than relying on intuition.
The tool was deliberately designed to be open-ended. We did not present a fixed set of pre-answered questions. Instead, we gave business teams the ability to explore the data themselves, form hypotheses, and test them on the spot. Pricing used it to identify segments for risk-based models. Product used it to evaluate contract structure changes. Legal used the gaming visualisations to build the case for tightening re-entry terms. The executive team tracked the impact of changes over time.
Results and Impact
| Metric | Value |
|---|---|
| Advanced pricing programme unlocked | Risk-based pricing worth £20m/year |
| Gaming patterns identified | Quantified for the first time across all channels |
| Contract changes implemented | Re-entry terms tightened based on survival evidence |
| Time to insight | Reduced from weeks to minutes |
| Stakeholder adoption | Pricing, product, legal, and executive teams all active users |
| Decision-making speed | Contract change approvals accelerated by providing shared evidence base |
The £20m figure represents the annual value of the advanced risk-based pricing programme that this work directly enabled. Without a quantified, behavioural view of customer survival, the pricing team could not segment risk accurately enough to move beyond flat-rate models. The survival analysis tool provided the missing foundation.
Key Takeaways
-
Exploratory tools beat fixed reports. Delivering a flexible, interactive application unlocked far more value than a predefined set of answers. The best insights came from questions we did not anticipate.
-
Visualisation made the invisible obvious. Gaming behaviour had been suspected for years, but overlaying cohort survival curves made the pattern so visually stark that it shifted the conversation from "is this happening?" to "how do we fix it?"
-
Shared evidence accelerates decisions. Getting pricing, product, legal, and executive teams onto the same interactive tool eliminated weeks of back-and-forth over competing interpretations. Contract changes that might have taken quarters to agree were approved in weeks.