5 minutes

Marketers and the Future of DMP Insights

Advertisers, agencies, and publishers are swimming in data. They have so many data points, from a variety of sources, that they are simply overwhelmed by it all. Website (cookie data), social data, CRM data, you name it, and they’ve likely got it. Sorting all of this data from various (often siloed) sources, in a timely and efficient manner is a near impossible human task.

We all know that the role of a marketer is to reach the right consumer, at the right time, with the right message. But to do this effectively, marketers are challenged with interpreting their mass amounts of data and uncovering actionable insight, at speed and scale.

Interpreting mass amounts of data is no easy feat.

As the demand for digital marketing and programmatic/real-time ad buying rises, marketers face more pressure than ever to target audiences faster, and with laser-precise, data-driven insights. We know that consumers will only respond to the messages that speak to their interests, passions, wants, and needs. And in the world of real-time bidding, technologies only have milliseconds to get that messaging right. And guess what? These messages cannot be crafted with broad categorization methods like demographics alone. Demographics as a stand alone are limiting and tell you nothing about what an individual is interested in, passionate about, or value.

To fill this gap, we have seen marketers seek more and more data resources. That’s why we see marketers not only trying to make sense of their first-party data but also second party data (from partners) and purchased third-party data. Can you understand why marketers are swimming in data? It’s a vicious cycle.

So again, we arrive at our original problem: how can marketers turn mass amounts of data into actionable insight, at speed and scale?

Are DMP’s the magic solution in the advertising ecosystem?

To better target potential consumers, many advertisers rely on Data Management Platforms (DMP’s) to collect their mass amounts of disparate audience data (including the first, second, and third-party data we spoke about) and interpret it. In short, DMP’s are cloud-based warehouses used to generate an audience segment(s) based on patterns and trends set within defined parameters. The goal, of course, is to deliver high-quality, accurate audience segments to marketers, and all other players in the advertising ecosystem, like DSP’s. When placed into action, these audience segments (generated by the DMP) should result in smarter optimized ads, efficient media spend, and less ad waste. But is this actually the case?

Marketers are sitting on a wealth of data, with a goldmine of potential insights to derive from that data. That’s why more and more companies are investing in DMP’s for their business and are hiring highly-qualified, expensive professionals to manage them. However, while DMP’s are used to extract insights, there is still a lot of wasted potential in these tools.

Here’s a quick DMP lesson: DMP’s operate on a “hypothesis” basis. DMP users must set conditions or a query to break down the data sources and form a specific audience segment they want. For a DMP to work properly (with speed and accuracy) and know what data to segment or pair, a DMP user must understand many factors including media, marketing, analytics and of course data. The DMP will then do its best to match data and form an actionable audience segment for the marketer to leverage.

For example, a marketer could leverage behavioural cookie data to build an audience of males in Nova Scotia, over 30 who browsed a car website on their mobile device. This audience can then be used for ad-buying, media placement, etc.

But marketers don’t know, what they don’t know.

But what does this marketer really know about this audience? What are their interests and passions, outside of cars, and how can they be determined? This is why, despite the integration of DMP’s, marketers still aren’t getting it right. While automated, there is still a human error in how DMP’s select which data to process and interpret.

Don’t get me wrong; there is incredible value in DMP’s but there is also an incredible opportunity present. Ultimately, the goal of leveraging a DMP is to provide a personalized consumer experience by relating to their interests and behaviours. But marketers are only grasping at the data that they are currently able to understand. Like I said, DMP’s operate on a hypothesis basis, contingent on the user’s understanding of the data.

We, as marketers, haven’t even scraped the surface of what is possible with DMP data. Marketers need a solution that looks beyond predetermined hypothesis and attributes. Instead, we need a solution that interprets unsupervised data and can discover the hidden relations and insights within audiences that marketers don’t yet know.

How do you foresee 2017 shaping up? How will DMP’s evolve?