MTN slashes prepaid customer churn using SAS BI prediction tools

MTN has successfully reduced prepaid customer churn by some 45% in strategic segments. It has also increased its period of customer retention by almost a year, thanks to the giant cellular network operator's utilisation of a leading business intelligence tool - SAS Enterprise Miner.

The data mining tool from SAS Institute, the leader in business intelligence (BI) solutions, has enabled MTN to easily identify high value customers who could be on the verge of leaving the network. Indeed, the predictive churn model has proved to be 84% accurate.

In addition to identifying potential lost customers - particularly the high value ones - MTN needed to devise ways of increasing their affinity to the network by targeting them with specially tailored marketing campaigns.

"The project with SAS addressed a very basic business need," explains Marcelle Georgiev, customer knowledge analyst, Customer Management, at MTN.

"Eighty percent of the South African cellular market is prepaid. This represents a huge number of customers with whom to interact.

"We therefore used SAS' Enterprise Miner to build models that would identify those customers that we really should be talking to, so that we could interact specifically and directly with them."

Mohan Namboodiri, senior consultant at SAS, says there are proven methods of promoting customer retention and reducing churn - but only when you know a lot about your customers.

"Prepaid cellphone customers, however, are anonymous," he says. "While a classic retention analysis exercise would focus on key demographics such as address and language, the only information we have on our prepaid customers is call patterns, usage data and - to a lesser extent - some buying behaviour," he explains.

MTN and SAS therefore built predictive churn models using data mining, enabling them to identify at-risk high-value customers. The specially devised marketing campaigns were then built.

"There was a significant difference in the results of the campaigns that used the modelling technique to decide which customers to target, as opposed to just randomly selecting customers," says Namboodiri.

For this project, the cellphone number was the key, and transactional behaviour, including numbers and duration of calls, as well as SMS patterns were studied.

"We were very impressed with SAS from a business point of view," says Najma Hay, MTN marketing consultant for prepaid campaigns. "The prepaid project was highly worthwhile, as our churn rate declined significantly and we've been able to retain targeted customers for far longer than we had originally anticipated."

MTN uses SAS business intelligence in several other areas as well.

"SAS is integral to all that we do in the area of retention, and provides real value in our marketing efforts," says Garth Brown, senior manager, CRM at MTN.