Given the rapid change in technology there is a rush to connect more data sets to the marketing planning cycle. Two recent articles:
- AC Nielsen is overlaying bank/financial/credit card information on to TV and online viewing behavior to provide richer (pun intended) information for selecting audiences. This is obviously done at the anonymous level, and likely through a third-party. So they know someone watches Top Chef and bought a new fly reel with a debit card.
- Facebook is taking the story of targeting based on purchase history on the road. By segmenting people based on brand/category usage it has determined the number of ads to show an individual in order to optimize ROI. (This is actually an extension of work done in other media, just new for Facebook.)
- There will be unlimited choices in what could be used to improve targeting; to the point that we will act more like conductors orchestrating the fragmented landscape than being an expert in any one of them.
- The sheer breadth of information implies we will have to trust the results, we won't know a priori what works or possibly even why as the effects of programmatic buying filter into other areas of marketing.
- With such a wide range of attributes the set of business rules that work to assemble the content to produce the most relevant 'ad' will get more challenging and shift to the realm of prediction or testing at a minimum.