en

Optimization of Non-Contractual Campaigns

Goals

  • Detection of active users with low probability of future purchase
  • Prioritization of incentives based on customer value

Data

  • User descriptions
  • Data from previous purchases
  • Product taxonomy

Analytics

  • Development of non-contractual models (no request for withdrawal)
  • Detection of users with a planned future purchase which could cannibalize the margin if included in campaigns
  • Detection of users with low probability of purchase that could be incentive campaign targets without risk of cannibalization

Business

  • Careful definition of both predictor variables and explanatory variables
  • Validation of the business sense of predictions obtained and the effects of explanatory variables

Results

  • Satisfactory prediction of users who would make a purchase in the coming weeks
  • Increased ROI from purchase incentive campaigns
  • Avoidance of cannibalization by anticipation of future purchases
  • Definition of appropriate incentives per user

This site uses cookies to ensure the best user experience. By continuing to navigate you are giving your consent to cookies being used according to our cookie policy.
Click the link for more information.