An abstract from OgilvyOne Worldwide on using CRISP-DM

Successful acquisition campaign starts with good targeting. Data Mining techniques such as Predictive Modeling help to select within a pool of external addresses those with higher probability for conversion. Data Mining nevertheless is much more than data modeling to identify qualified prospects. It is an integral part of the CRM process, from definition of goals to implementation.

CRISP-DM, which stands for CRoss Industry Standard for Data Mining, structures the whole Data Mining project through a step-by-step approach. The attached presentation takes you through the different steps in the CRISP-DM process, how much time each of the phases take and demonstrates how CRISP-DM was successfully applied. It is especially useful for large Data Mining projects, making them more manageable, faster, cheaper and more reliable.

The presentation also holds great Data Mining tips such as:
(Note: the above presentation is in German.)

• identify a reference group to refine your target audience
• define Key Success Indicators early on
• clarify project restrictions such as persons who can't be contacted
• determine address selection volume up front
• pre-define monitoring objectives.