Real-Time Matters Part 2: Dynamic Audiences


In part one of “Real-Time Matters” we referenced Chris O’Hara’s recent AdExchanger article about real-time marketing and talked about how is helping advertisers source the freshest data possible for their advertising campaigns, from website and physical location visits to contextual categories and keywords searched.

We also talked about the benefits of real-time unstructured data elements as opposed to third-party segments. O’Hara brings up another great point in his article that user’s intent is fluid. No user is strictly a “pizza” or “auto” intender; User intent flows across devices and is not static. However, O’Hara says a solution to matching users across devices and intent is “dynamic segmentation,” the act of updating third-party segments at the moment a user’s intent changes.’s solution is to eliminate third-party segments altogether. Segments are opaque, as advertisers don’t know what types of users are within segments, let alone the data elements that placed the users in the segment. Rather than updating static third-party segments, builds dynamic audiences on the fly using unstructured data, targets users who have shown interest or intent in an advertiser’s product service, and optimizes to those users based upon the highest performing data elements.

The ability to execute effective real-time marketing relies on an advertiser’s ability to build the correct audience. With, advertisers can build their own Dynamic Audiences using up to 10,000 targeting elements as opposed to the typical 10-20 targeting elements of a segment-based DSP.

The platform then uses those data elements to optimize towards the factors that are performing the best and move away from the factors that are underperforming. The key to real-time marketing with unstructured data is finding the data and users that are performing well and optimizing towards those elements in order to create a highly targeted, dynamic audience.