Using SAS to determine clients' lifetime value

Companies within the financial services sector are trying to align their objectives of extending customer value with their intermediaries' focus on signing new business - any new businesses - as much and as quickly as possible.

"The company and the intermediary have different ideas as to the measures of value," explains Nigel Wigram, consultant to Old Mutual, South Africa, a major life assurance and financial services group with more than four million clients in SA.

"The company makes its money through the turn on money under management, through risk and mortality profits, and through expense charges over a number of years while the salesperson is primarily interested in new business. Companies are slowly trying to align these objectives by spreading the initial commission and by paying trail, maintenance or service commission."

Wigram says evaluating lifetime value is about looking for the real factors that determine the potential value that an organisation will extract from its client base over a period of not less than five years.

He believes lifetime value goes beyond the value of the contracts currently in force for a client, and the probability of retention of those contracts. It also involves the potential for future business from the client - the likelihood of the client purchasing new products and retaining them; and even intangible factors such as the value of the client as an advocate and influencer of future business. All that has to be weighed against the marginal or optional costs involved in retaining that client.

"In effect, this means giving most attention to those people who can provide the best value into the future - and trying to avoid those providing little probable value. However, the biggest profit comes from recognising, enhancing and exploiting identified value increasing factors for each client: factors such as a change in life circumstances," he adds.

It also means identifying potential value-decreasing factors for clients - and then taking steps to minimise the impact on the client, and therefore the company.

"There is a danger that short-term profit considerations in relation to high risk policies could lead us to gather a few more premiums from a doomed policy - and merely increase the loss experienced by the client.

"Every lifetime value equation must include an advocacy factor to remind us that if we antagonise these clients, they are likely to tell their friends and this could lose us future business. Certainly there is a reason to disengage gracefully from some of these clients. However, in most cases keeping the policy going for a few months by handling the immediate threats and problems substantially increases the probability that it will survive for many years in the future.

"Indeed, interventions to lower the risk of loss in marginally profitable segments often have a multiplier value well in excess of the immediate effect on the client," he adds.

Old Mutual's operational systems form the data sources which feed the marketing data warehouse that is mined to determine clients' lifetime value. This warehouse contains current and derived data from a wide range of sources - intermediaries and client contact via the Internet and the call centre - as well as a history file of all the events or changes that have been tracked by the company.

Base SAS is then used to create a number of marts which are accessed via NT servers. These marts are then used for a variety of purposes including campaign origination - using a home built tool called Segmentor, and analysis and mining using SAS Enterprise Miner and SAS-based Futrix tools.

"Careful analysis of this data is helping to better determine the potential lifetime value of our clients - and to identify and act those factors which will ensure we extract the greatest possible value from the client," Wigram concludes.