Predictive Targeting – How DMPs can give you novel information about your users



Ad Networks and their media buying teams are always stuck in a maze of perpetual uncertainty. Users are tracked and barely understood until entering downloaded apps, only to disappear into a puff of smoke. The post-install metrics that are so coveted by UA media buyers are seldom shared, which is reasonable given that this data is an incredibly valuable commodity for their advertising partners.

With User Acquisition becoming such a valuable tool for apps, and the skill-set required becoming so dynamic, this siloed approach to data management is restrictive. For a Performance Marketing Agency with DMP capabilities like BidMotion, the challenge of withheld data is met creatively rather than overcome. We leverage what our partners don’t have. Namely; the experience of running thousands of campaigns with hundreds of apps and the cumulative learning engendered by this experience.

The targeting parameters available to all networks (e.g. Device type, version, OS, ISP etc.) are the most common metrics to optimize across. Yet many networks are not accruing the data on these metrics, leading to a careless loss of competitive advantage. For a more sophisticated network, optimizing across metrics like Publisher ID and Network ID can be a very useful way to exclude poorly performing inventory and determine what sources targeted audiences are most likely to use; an extremely valuable piece of information.

Not only can these models help determine general audience segments, but also they can be useful in predicting personalized data points like age, gender, interests, and income level. For this latter data point, it is often crucial to have some post-install information. Even without this info, however, predictive modelling is still effective, albeit it at a lower confidence level, something which leaves uncertainty but is considerably better than relying on chance alone.

In a network’s ideal world post-install data would be freely available, allowing media buyers to know exactly who the most engaged users are and from what traffic sources they were found. As it stands, networks must use the information available to them, as well as their cunning and intuition, in order to determine where and when to buy inventory. If traffic sources aren’t taking advantage of the data they have accumulated, it’s time to find a partner who does.