A European financial service provider (bank accounts, loans, mortgages, insurance, investments) carried out some relatively simple analysis to look at customer profitability. To do this, the company calculated the current value (CV) and net present value (NPV) from predictable income (for example, from committed revenue streams such as loan and mortgage payments planned) and NPV of less predictable income from cross-selling. It attempted to build in a contribution from the 'recommended' business from that customer, but for this company it proved too unreliable.

It then carried out a decile analysis combining CV and NPV of current customers, using actual and committed income. This showed that 77 per cent of the profit came from 20 per cent of the base. It used the results of some carefully constructed cross-selling tests to develop propensity models. These models were tested and adjusted to be, at worst, 68 per cent predictive. Then, based on these CV and NPV deciles, it plotted where the future value lay. The analysis chart showed that although the top decile had probably achieved its peak (older customers in this case), deciles 2 and 3 had a great deal of potential, with deciles 7 and 8 also containing some potential high-value segments (younger customers, still early on their value curve for this company). This was useful in that it helped it value the customer base, determine retention, development and 'exit encouragement' programmes to create more value, and develop segment profiles usable for acquisition, retention and development planning.

Application of this knowledge has clearly created value for the company, as can be measured via the database. Possibly more importantly, this work led to a step-change in the understanding this company has of how to manage customers.