"Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by John Lee, executive vice president at Merkle.
Challenged by an already short digital marketer attention span, the data-management platform (DMP) is already running the risk of going from today’s front-page news to the “Where are they now?” column all too quickly.
It’s not because there isn’t a lot of value to be had for marketers interested in maximizing their digital media and site personalization dollars. Today, advertisers can onboard valuable first-party segments and develop flexible online models and audiences that can be reached across multiple platforms. The audience creation, campaign management, syndication and analytic insight functionalities are very powerful and are driving measurable lift for some data-driven marketers.
In spite of this, several risk factors obscure the value and contribute to the potential premature demise of this once promising star.
Adoption Without Purpose
I’ve been amazed to see the number of advertisers who implemented a DMP without really understanding what it does or how it will add value. The DMPs and agencies don’t seem interested in bringing clarity. If they could lay out a simple set of core use cases that show improved performance with increased digital media and site personalization, then advertisers would be able to align their goals accordingly.
By the same token, advertisers are not clearly defining what realistic problems they are trying to solve and therefore have no way of knowing if their DMP-run programs are really working. Instead the conversation turns to things like onboarding, accessing third-party data lists and better retargeting, which are features that are not really part of a DMP’s unique value. As advertisers learn and understand this more, they will justifiably begin to question that value, and they’ll wonder why they need to pay for a DMP when their demand-side platform (DSP) or site optimization tool can – with a little elbow grease – already do these things.
Another source of risk is self-inflicted confusion in the ad tech marketplace around what a DMP really is and who really has one. There are a few very real business needs that the DMP solves, but it doesn’t do everything. For instance, it doesn’t connect to your paid social platforms, determine who your highest-value customers are or fix your data integration problems. Since it isn’t truly omnichannel, it also can’t manage the master consumer ID of record. Yet snake-oil claims abound from the DMP suppliers.
Worse, suddenly everyone is a DMP. Every DSP is a DMP; tag-management solutions are DMPs; and database providers are DMPs (or cloud operating systems with overlapping DMP capabilities). In reality, there are no more than four or five meaningful buy-side DMPs that meet basic requirements. This confusion makes the advertiser’s head spin. And as reality sets in, so do disappointment and a lack of real adoption. I wish the major ad tech players could be satisfied with doing one or two things really well.
The Cookie Has Crumbled
More than any of the above factors, the cookieless future is also starting to cast a shadow on a DMP landscape that has placed all of its chips on the third-party cookie.
Yes, the major DMPs are working on incorporating cross-device tracking through nascent probabilistic partners, but there are just too many questions around the accuracy of the analytics for it to be meaningful. PII-to-cookie onboarding is a useful tool for analytics, but its low match rates really limit its value in targeting.
I'm seeing more addressable media plans now, where the mix is shifting heavily toward use of “person” vs. cookie IDs in Facebook and Twitter and away from desktop real-time-bid. The current announcement from Facebook on the revamped Atlas that will have a person-based ad network and bidder will accelerate this phenomenon tremendously. I can’t imagine that Facebook won’t build its own DMP inside its walled garden. This factor – beyond any of those listed above – could spell the premature end of the DMP era.
The DMPs should consider major changes in their approach and begin to partner on person-level data (they currently can’t handle the PII). The DMPs that are embedded within broader marketing cloud solutions like Adobe should take a hard look at how the DMP and the traditional campaign management tools fit together as a single tool. As more dollars flow into person-based media, the ability to serve both the person and cookie sides of the equation in one toolset will become a powerful value proposition – especially when partnered with the CRM database providers who can handle the PII.
The DMP can be a valuable tool in an overall process that involves the melding of first- and third-party data to inform better targeting with digital media and personalization. But before this promising technology can really take root, agencies and DMPs must strip away the confusion and pivot toward the person-based future.