Case Study · Insurance · Short-Term · Multi-Brand
£682m of premium. 1.4 million customers. Five brands.
The data said they were not one customer base.
A multi-brand short-term insurer. £682m of Gross Written Premium. Five brands built for different needs, price points and life stages. A performing asset, and investors who wanted to know where the value creation levers were. A customer base that tells a more complicated story about where they actually sit.
Retention said one thing. Margin said another.
When every customer was scored using the Keystone IQ Revenue Quality Architecture™ and tracked by brand and by cohort, three numbers emerged that do not appear in portfolio reporting.
£47m
Annual margin the existing Brand C customer base supports on its own, after observed churn
93%
Of customers who left one brand left the Group altogether, rather than moving to another proposition
+£5.8m
Identified annualised margin from the existing base, no additional customers assumed
None of this was visible in customer counts, premium income or margin, which is all the internal reporting produced.
What the customer base actually showed
93% of customers who left one brand left the Group, so five brands were running as five acquisition funnels rather than one compounding base. The cohort with the second-highest three-year retention in Brand A, at 44.9%, carried a margin of minus 6.3%. The same customer relationship produced margins 83 points apart depending on which brand it sat in. And Brand C's existing base, 66,006 customers retained from an opening 183,860, supported £97m of Gross Written Premium and £47m of annual margin. See how the analysis is built.
None of this appears in standard reporting. All of it changes where you look for value.
The brand story
Five brands were built for different customer needs, price points and life stages, on the assumption that a customer who outgrows one proposition moves to another and the Group keeps the relationship. Only 7% of customers leaving a brand went on to take another. The other 93% left altogether. The answer is not an indiscriminate cross-sell campaign: 259,840 customers held motor cover with the Group and no home policy, and 42,560 held home cover with the Group and motor cover somewhere else. Those are qualification pools, and they are only visible once the base is rebuilt across brands rather than reported inside them.
The retention story
Reported retention did not identify where value sat. Once claims, price and servicing economics were applied to the same retained relationship the ranking inverted: in Brand A, the cohort with the second-highest three-year retention at 44.9% carried a margin of minus 6.3%, while the cohort with the lowest retention of the four carried the highest margin at 29.5%. Churn timing moved as well, running from months 1 to 3 for one cohort to month 24 onwards for another, so an annual renewal programme arrives after the decision has already been made. See revenue quality.
The floor story
The £682m headline is real. What it does not show is the margin the existing base supports on its own. For Brand C, an opening base of 183,860 customers retained 66,006, and those customers supported £97m of Gross Written Premium and £47m of annual margin, 46% of estimated brand GWP. That £47m is a protectable earnings floor, built on observed cohort retention and excluding all future acquisition. Four changes, none of them requiring a single additional customer, lift it by £5.8m, roughly 12% of the baseline. Related: the Health Check and the Value Creation Map.
Download the full insurance case study
The complete read, including the eight customer relationships, the brand and cohort margin matrix, and the four sized value creation levers.
Download the full case studyFAQs
It shows which customer relationships carry durable margin and which do not. Keystone IQ rebuilds the book from anonymised policy and transaction data at customer level, then measures retention, claims economics, tenure, replacement dependency and next-need eligibility by cohort and by brand. In this engagement it showed that 93% of customers who left one brand left the Group altogether, and that the same customer relationship produced margins 83 points apart across three brands.
Because a retained customer can be an unprofitable one. In this book, the cohort with the second-highest three-year retention in Brand A, at 44.9%, carried a margin of minus 6.3%. The cohort with the lowest retention of the four carried the highest margin at 29.5%. Retention counts customers. Retained contribution margin counts value, and the two can point in opposite directions.
It is the annual margin the existing customer base supports after observed churn, before any future acquisition. For Brand C in this engagement it was £47m, against £97m of retained Gross Written Premium and 66,006 retained customers from an opening base of 183,860. It is a protectable earnings floor rather than a forecast, and it is the number a value creation plan should be judged against.
Not automatically. In this Group, five brands designed for different needs, price points and life stages were largely operating as five separate acquisition and retention portfolios. Only 7% of customers leaving one brand took another. The value in a multi-brand structure comes from routing customers to the proposition where their relationship carries the most durable margin, which requires eligibility rules, not cross-sell campaigns.
Yes. Keystone IQ requires no names, no contact details and nothing that identifies a person. The analysis runs on anonymised, customer-level transaction and policy data. Identity is withheld throughout and all £ figures in this case study are approximate conversions from local currency.
See exit readiness and the full set of customer base analysis case studies.
Identity withheld. Data anonymised. All £ figures are approximate conversions from local currency. The Brand C retained-base scenario applies observed cohort churn to the starting base and shows the resulting customer, GWP and margin floor before future acquisition. Retention rates are conditional, year-specific retained proportions, not cumulative cohort survival. Keystone IQ Revenue Quality Architecture™.
See what your customer base is really made of
Days not weeks. Fixed scope. No PII required.
Get in Touch