With today’s announcement that marketing tech platform IgnitionOne is acquiring Knotice, the ranks of stand-alone data management platforms (DMP’s) are dwindling to a mere handful. Oracle recently picked off market leader BlueKai, and before that Neustar merged with Aggregate Knowledge . . . and before that Hearst bought Red Aril and Adobe bought Demdex . . . and digital marketers who are not deep into ad tech are left secretly wondering what all the fuss is about.
So what is a DMP, and why does everybody want one?
As I see it, DMP’s do six things that are very important to marketers. They:
- take in data from many different sources that ordinarily do not connect, speak the same language, or sit on the same server
- combine this data into a single useful repository of information where like items are formatted and labeled in the same way, so the DMP can then . . .
- tie that data to an identity (either an anonymous or a known person or a particular browser), often by “synching” cookies from different providers who have agreed to do so or by using a company’s own customer data
- build up millions and millions of such profiles in the cloud
- enrich these (incomplete) profiles if possible with information from outside data providers, at a cost, to add dimensions such as geography or gender
- help marketers determine which among the millions of user profiles in the database might be good targets for their purposes
A traditional DMP stops here. The next step is applying analytics, say, for predictive modeling, and executing the program, perhaps through bidding for impressions on advertising exchanges, search marketing platforms, or social networks. It will surprise no one to learn that DMP’s in recent years have been adding these “next steps” at a rapid clip, even as traditional demand-side platforms (DSP’s) such as Turn and DataXu, which began as tools to execute ad campaigns, have been building out their own DMP capabilities.
No two DMP’s are exactly alike. They started in different places six or seven years ago, and the dreamers who started them didn’t know exactly where the market was going. We still don’t. BlueKai began as a data exchange, where marketers could share and synch anonymous cookie tracking data to improve reach and targeting. Demdex focused on audience segmentation for publishers who sold advertising space. Aggregate Knowledge hung its hat on campaign analytics and privacy controls.
Likewise, none of the combined DMP-plus-acquirer brands will look the same. A DMP alone does not make a digital marketing hub. Each of the providers calling themselves a Digital Marketing Cloud/Hub/Suite, with or without a DMP, lacks something that someone else has. That’s why I think the DM Hub story is rapidly turning from a land grab into urban development. Who is going to execute the vision best?
So: Why does everybody want a DMP?
It’s not glamorous, but I see the DMP as the “middleware” layer of the digital marketing system. (I’m borrowing this analogy from David Raab and Scott Brinker, and AgilOne’s Dominique Levin convinced me it works.) In the same way middleware connects different systems, DMP’s have emerged as the layer that brings together (1) anonymous information across different channels, and (2) non-anonymous information from company databases. Without something like a DMP, this data would still exist, but it could only be used in isolation, channel by channel. (This is how it’s still used by most marketers.)
DMP’s are not the only potential connectors. Tag management systems (TMS) and some analytics solutions perform many of the same functions. (TMS’s are getting their own love from investors recently.) But it’s becoming clear that any provider with aspirations to a digital marketing Hub needs to act like any good marketer and start connecting, fast.