...
define which metric to optimize (average basket, conversion vs impression...) and score the entire database of Weborama against this parameter and then arbitrate, when selecting a target, between impact (% of internet users exposed) and expected performance (performance index)
extend a target audience by targeting customers in Weborama’s database that have a very similar profile than the customers who converted to a specific campaign (= look-alike audience)
understand which parameters drive the most value for a campaign (predictive segmentation tree).
