Situation
KPN offers telecommunication services to both consumers and businesses.  KPN's core business comprises telephony and data services through KPN’s fixed network in the Netherlands, mobile telecom services in Germany, the Netherlands and Belgium and data services in Western Europe. KPN is market leader in all major segments of the Dutch telecom market. Through E-Plus in Germany and BASE in Belgium, KPN has number-three positions in the mobile markets of these countries.

Challenge
The telecom market is a very competitive market. Whereas a number of years ago the emphasis was mainly on finding new customers to increase market share, the current emphasis is on keeping existing customers. The average change percentage of customers in the mobile telecom market is approximately 20-25 per cent. KPN Mobile believes predicting customer behavior and needs is an important method to further improve the effectiveness of its direct marketing campaigns.

To contract more customers, increase the cross-sell ratio and bring down the churn, and at the same time considerably improve the marketing process, KPN Mobile specified the following project objectives:

• Efficiency improvement of the DM process by 15 per cent. Considerable reduction of the set up and execution time of campaigns.  
• Effectiveness increase of the DM process by 10 per cent. The result of the actions needed to be 10 per cent higher.

Solution
KPN Mobile’s starting point was the creation of a more efficient process to build models.  Using predictive analytical software the target group could be defined more quickly and more efficiently after consultation with the marketer and the campaign manager, resulting in a substantial reduction of the execution time.

Implementation
KPN Mobile has divided its database in three segments:  business market, consumer market post-paid and consumer market pre-paid. The implementation project was aimed at creating an analytical environment for the post-paid and prepaid campaigns.
 
KPN Mobile also uses the solution for the telemarketing and direct mail channels where the costs per customer contact are high. And the software is used for the development and management of customer segmentations. Accurate predictions are then made of the expected customer response, so that marketing campaigns with the right message can be aimed at the correct target group.

As part of KPN Mobile’s customer value management strategy the organization aims at keeping valuable customers and to develop customers.  By using predictive analytics software to predict which customers might purchase more products the profitability of customers can be corrected.  In addition, the tools allow KPN Mobile to easily identify risk groups with a high churn risk on time in the base and to react proactively through actions.

Result
The use of predictive analytics software has, in addition to an acceleration in the marketing processes, resulted in a substantial cost reduction of a number of campaigns and therefore an increase in the profitability per customer. The estimates for the first year amount to cost cuts in excess of one million euro for the top-three campaigns where telemarketing was used. 

• The campaign efficiency improvement goal was set at 15 per cent. Based on the measurement of the turnaround time of campaigns carried out with and without PredictiveMarketing, the improvement has resulted in a halving of the time that is required to carry out campaign selections. Whereas the execution of a target group selection used to take a day this has been reduced to just a couple of hours and in some cases to a press of a button for the lifecycle models. 
• The effectiveness goal – an improvement of 10 per cent – was achieved for a number of important campaigns where the expensive telemarketing channel was used. By using cheaper channels, 10 per cent less costs were incurred and the result was the same. 
• With the same marketing budget strongly improved results have been achieved.