Will This CRM Program Pay Out Soon?

 

ABC Bank was considering a new CRM program to attract and retain customers over a 3-year period. The prospect universe was 10 million households and research suggested the offer was likely to attract 4% each quarter at an initial cost of $100 per customer.

After the first quarter, a newly acquired customer could be expected to generate a quarterly profit of $25 each, assuming he or she didn't defect. This contribution level could be expected to remain steady so long as the customer was retained. The bad news was that historic data indicated 20% would defect in the first quarter. The good news was that 5% could be expected to immediately evolve into "Highly Developed" customers, generating $75 in quarterly profits for each successive quarter.

If a customer could be retained into a second quarter, ABC could reduce the likely defection rate to 10% and increase the evolution to Highly Developed to 30%.

95% of customers who became Highly Developed stayed that way, with only 1% defecting.

The program start-up costs were estimated at $60 million, with another $15 million per quarter in operating costs. The bank faced two questions: (1) Could they generate a profit and when? (2) Just how far could this program penetrate the prospect universe over time?

A simple Markov chain template answers both questions. Markov chains are useful to look at how customer populations might change over time. They can be used to examine changes in market shares, product trials, distribution patterns, or just about any other variable that might move from one condition to another in the face of market fluctuations. (See Introduction to Forecasting with Markov Chains.)

We laid out this information to create the Transition Matrix below to show how consumers would move between segments.

Then we looked at how the consumer universe would change over the 36 quarters and hat profits could be expected per quarter.

 

Finally, we looked at the cumulative profit to see if we could make money in the 3-year timeframe and examined prospect universe penetration.

 

In less than 15 minutes, we discovered the program could produce a profitable outcome in under three years and should penetrate about 30% of the prospect universe during that period, increasing to 40% by the end of Year 5.

Most important, the interactive tool we built to answer these questions can easily be used to experiment with different assumptions about customer mobility between segments, program costs, and other variables.

Want to download our sample worksheet? Click here for the CRM Markov tool.