Extremely precise audience modelling opens vast new possibilities for digital marketers.
DMP – Data Management Platform – is a special data warehouse used for classifying, analyzing and processing audience data collected from innumerable sources. The top level data classes in a DMP are:
- First party data: Collected or created by a company itself, owned data. Data sources include visitor data from web pages or social media, items viewed or purchased in online stores, as well as customer data from marketing automation, CRM or ERP systems. Data is organized using a self-defined hierarchic taxonomy of classes.
- Second party data: Another company’s first party data you have permission to use.
- Third party data: Anonymized data collected by a large number of DMP data partners (business registers, media, credit card companies, online stores etc.) from millions of websites. Contains information on media usage, hobbies, interests, occupations, purchases intended or completed, demographics etc., classified into a deep taxonomy of tens of thousands of nodes.
Anonymization of all data in DMP enables audience building over class boundaries. For example:
- C-level managers of all Finnish companies employing more than 500, with your own existing customers, prospects and leads excluded. The audience is updated constantly with members flowing in and out, meaning the data is always up-to-date and no “become a customer” ads are shown to members already in the sales pipeline.
- All property owners over 40 years of age with Android devices, who have purchased your product A but not product B.
Additionally, look-alike modelling can be used in audience building, e.g. to identify the most likely buyers from third party data using your own existing customers as seed. This may be the easiest way to target the competitor’s customers.
Programmatic advertising finds its audience, wherever they are
Data collected and classified with care can be used in many ways. DMP’s best pal, Demand Side Platform or DSP, is used for programmatic buying of display, social or search advertising. The central benefit of programmatic buying is that the ad is not placed on pre-defined locations. Instead, it always finds its audience wherever they surf. When a member of the targeted audience enters e.g. a news site where programmatic ads are shown, DSP participates in an auction. The highest bidder’s ad will be shown.
The use of DMP as an aid in programmatic buying enables targeting ads to precisely defined audiences, which in turn saves money. For a big global advertiser, the savings may be vast. DMP-based targeting also enables you to switch ad versions on the fly or to terminate the campaign after a verified purchase. Programmatic cross-selling is also a common scenario. Even if it is now becoming more and more common that every DSP has a built-in mini DMP, running an independent DMP of your own enables platform agnostic use of any number of DSPs and enhances first party data gathering and analysis.
For the first time it is now possible to automate the whole advertising and business communications funnel with all its channels by using a DMP in tandem with an integrated marketing automation system. This is a historic change offering unique, new possibilities for even the most modern of marketers. Use of real-time data opens unforeseen possibilities for dialogue, creative storytelling and meeting the customer.