• by Christopher Hansen , April 4, 2013

Data management platforms, or DMPs, seem to be the hottest tool in online marketing today. As the industry learns of the advantages of big data, marketers and publishers are under the impression that a fancy piece of technology will help them gain control over all their data sources. DMPs are helpful in certain circumstances, but most marketers are signing contracts without truly planning out what they will do with the technology once they have it.

The truth is that DMPs are an expensive splurge, and one that is not necessary for all marketers. Before licensing a DMP, marketers need to ask themselves whether they will actually get value out of the technology and what actionable results they expect from the integration.

The DMP pitch typically offers marketers five different benefits pertaining to data: centralization, security, portability, aggregation and analytics. Each of these functions sounds appealing, especially to a company that is concerned about its “big data’” strategy. But the bigger question is how each of these services will be utilized once the DMP is in place.

The value of big data is taking an abundance of unstructured information and sorting it out to produce insights that lead to better marketing initiatives. But before that happens, the enterprise needs to define what it is they want, or else they will be sitting with a lot of information that can’t be used. A DMP will not answer all data questions at the click of a button -- even marketers with interesting first-party data still need help leveraging that data to find new audiences for ad campaigns.

Targeting based on big data is not an art -- it’s a science, and unless you have a full-fledged data scientist on your staff, you probably won’t be able to leverage a DMP to its full potential. Few brands and agencies have an internal expert who understands data enough and has the appropriate training to use all of the DMP capabilities as intended.

Let’s compare the DMP to the concept of gym memberships. Gyms operate on the assumption that many of their members will pay a monthly fee and never set foot inside, benefiting from consumers’ intentions on making a positive change. Gyms can proceed to over-sell the number of memberships because there is very little chance the gym will become overcrowded. When a consumer does decide to take advantage of the facilities, a few will ride an exercise bike, take a class, or use the treadmill, but to actually accomplish any significant goals and visible gains, they will likely need to hire an expert -- a personal trainer -- at an extra cost. The same goes for the DMP. It provides marketers with the equipment, but they still need to know how to operate the tool for it to be effective.

Existing technologies -- including ad networks, audience targeting platforms and DSPs -- have experts in house that have built algorithms specifically designed to help their clients find a target audience across the Web. Many have data management capabilities integrated into their platforms already. More often than not, data management, analysis and data science services are baked into the pricing model too. This is the heavy lifting that results in actionable insights.

Furthermore, this data expertise can be applied immediately to display buying, whereas self-service DMPs disaggregate the data from the media, eliminating the efficiencies that marketers get from one platform managing both. To make the DMP data actionable, marketers first need to gather the insights and then port out the data to other third parties.

DMPs promise security, exploiting trust issues concerning data leakage in the ad technology space. Marketers are told their data isn’t safe unless it is centralized. But this claim falls apart when marketers want to activate their data with media, because that’s only possible by sharing it with publishers, networks and DSPs. Porting out data from the DMP to these third parties, as opposed to directly, doesn’t make the data any more secure than it was in the first place.

DMPs offer value, but only if the marketer knows what they’re doing and has a clear intent for the technology. Licensing a DMP can be a costly decision, and one that should not be made based solely on ad tech chatter.

Before taking the plunge on a big piece of data technology, marketers need to ask whether the tech justifies the effort and price. Implementing a DMP solely to centralize data in one place should not be the end game. In short, figure out what you’re getting, as well as what you may be giving up.