Recently the IAB and Winterberry released “The Data Management Platform: Foundation for Right-Time Customer Engagement,” a whitepaper outlining the growth, possible evolution and hypothetical use cases for Data Management Platforms (DMPs). The report is insightful and well-organized, and overall, provides a pretty solid picture of where DMP technology is headed. However, I believe there are additional key points to be addressed.
Let’s first recognize that DMPs have always been around – even before they were called DMPs. The reason they’re attracting the spotlight now is that big data’s day is drawing near. To that point, what's happening on the frontier of big data is far more relevant for DMPs as well. The industry at large is starting to understand that DMPs can only do so much when they work in isolation. Far better analytics and insights can be drawn from data that's beyond the scope of a single DMP. By integrating more third-party data with first-party data, we can generate better results on most fronts.
To address another point, predictive analytics represent one of the biggest opportunities for big data in general. As an industry, we have barely scratched the surface of its potential, but the emerging innovation around predictive analytics is likely to drive the better part of DMP technology evolution. As a result, we will see more momentum towards making insights actionable. It's one thing to analyze historical data; it’s quite another to deliver a recommendation that yields better business. We’ll see more innovation on this front in the next 12-18 months.
The need for specific analytics and insights will drive more innovation than the size of the data itself. Big data doesn’t always need to be “big.” More often, it needs to be synthesized, particularly for marketers, who are looking for specific points within the data to identify, reach and activate their audiences. The data doesn’t always have to be bigger for them - there’s already so much they don’t use. For them, the data actually needs to be more specific – although it may still need to be drawn from multiple sources in addition to their own.
On a related note, we will likely see more "open-source" thinking about data in general. Conventionally, most publishers and advertisers have held on to data that just hasn't generated enough value in their businesses. By building partnerships with leading technology innovation companies, they may gain capabilities that will help them better leverage their own data gold-mine.
Big data is making it easier to integrate across various services, sources and forms of data. This enables, with some planning, a much better opportunity to deliver better insights that take into account a variety of signals, far beyond what one system would offer. As an example, we might see the effect of weather on online shopping behavior, or how a blackout in one area affects the performance of certain ad categories.
The technology behind big data is making incredible advancements. The computational power accessible to any company today is simply phenomenal compared to what was available just five years ago. Previously, a deep study of analytics required custom or dedicated hardware and software solutions that were costly and only available through large solutions providers. Today commodity machines are making it much easier to do more than we ever thought possible. Data is more abundant and more accessible than ever. For that reason, DMPs will certainly have a permanent place in our online advertising ecosystem, and in our data-obsessed industry, they’re poised to become an indispensable tool for marketing and advertising professionals.
This article was co-written with Erik Pavelka, COO, Martini Media