Although Data management platforms (DMPs) are used by the lionshare of America’s largest companies, there is still much confusion regarding the technical requirements for a DMP to truly do its job. Recently, HP began a global expansion of its use of DMP products with data partner BlueKai. The Advertising Technology Review spoke with Wayne Thorsen, SVP, Strategic Partnerships at BlueKai about how DMPs ought to manage their client’s data and what CMOs and CIOs need to keep in mind when evaluating a DMP. Not surprisingly, he’s quite keen on his own company’s offerings, but we won’t hold that against him.
The line between demand-side platforms (DSPs) and DMPs has been muddied a bit as both types of platforms gain popularity. What does a good DMP do for its clients?
A quality DMP not only manages, organizes and segments your data in a meaningful way, but it allows you to take action on all of your data assets, across all of your media channels. Boiling this down into functionalities, a worth-while DMP will allow easy data ingestion, for example. This means that a DMP should allow you to bring all of their audience and ad data together in one easy-to-use platform. It should be able to pull data from your site and any affiliated partners, as well as data warehouses – customers, purchases, location, demographics, and meta-data such as account worth etc – into a centralized taxonomy.
A DMP should offer media-agnostic data portability in order to maintain maximum data control.
A DMP should offer media-agnostic data portability in order to maintain maximum data control.When you work with one or more DSPs or SSPs, as well as multiple networks or exchanges, your DMP should be able to push and pull data to and from all of these sources, and then analyze this data in one place to get cross-channel and cross-media insights. The most efficient way to support data portability is to work with a vendor that has pre-integrated a vast list of media partners through server-to-server transfer technology and offers direct integration of first party data into the publisher’s ad server. DMPs should also be fully transactable, allowing you to get granular audience intelligence and then optimize inventory allocation with one click for analytics, sharing features, and third-party data purchases. A DMP should include access to third-party intent data culled from millions of users across the web, because this data is critical for building new, lucrative audiences, setting inventory pricing, and analyzing audience effectiveness. It should also offer clear insights into how each and every audience segment is performing at any given time are critical, and every DMP should offer robust data analytics. A DMP should allow you to measure how specific audiences interact across multiple properties, and generate reports showing all data attributes associated with a specific audience form all available data sources.
Publishers are at the core of the advertising technology industry, but much of the focus has been on advertisers. What do publishers need to keep in mind when selecting a DMP?
if your DMP limits your integrations, your data ingestion, or access to your data assets you will not be able to monetize and drive yield from your site audiences. An effective DMP will help centralize disparate audiences into one pool, allowing publishers to gain a much better understanding of their audience assets across all properties. BlueKai, for example, has the deepest industry integrations and partnerships, including the tightest integrations with DFP, integrations into all four Portals, top 30 ad nets, DSPs,and trading desks for data acquisition. In addition, a DMP should provide easy or built-in access to vast amounts of third party data. This is necessary to provide rich analytics and/or insights reporting.
How has the publisher’s relationship to data changed in the past few years and how do you see that relationship changing?
In 2005, experts reported that the amount of digital information in the world was doubling every 1100 days. In 2007, that time had shortened to eleven months. A recent study by IBM predicted that in the next few years the amount of information will double every eleven hours.
This data overload has made it difficult for publishers to separate signal from noise, or meaningful customer attributes from the whirlpool of customer information out there.
This data overload has made it difficult for publishers to separate signal from noise, or meaningful customer attributes from the whirlpool of customer information out there.This complex relationship is addressable and containable within the right system. The key, is to manage your data in a non-static way, and not try to fit a square peg into a round hole. Publishers need to evolve as their data evolves, and that means choosing an agile DMP solution.
Many DSPs offer DMP-like services, what are the core differences and benefits of dealing with one or the other?
While there’s room for DSPs in the marketplace, DMPs offer complementary and endemic capabilities, and most importantly are media-agnostic. If you use a media partner such as a DSP or another network as your DMP, your data assets may be stranded, with that media partner, or difficult to use with competitive media partners. We call this a data tomb. Your data is locked in there forever. We believe in portability, and having a DMP that facilities this portability to any partner so clients can work with whoever they wish. Even though your DMP should be media-agnostic so that the platform isn’t incentivized to get you to spend more, all your decisions should be data-driven. Many platforms say they have the most access to data, but the reality is most have to buy it, and they have a small percentage, usually ten percent. It is important to note what type of access a DSP or DMP has to a data exchange in terms of source, type, and amount.