Programmatic Fuels the Audience Engine

August 1, 2014

How PMB is empowering marketers to connect with the right consumer every time

By Justin Evans

Programmatic media buying (PMB) appears extraordinarily fragmented compared to buying digital media directly from a publisher. That’s because your campaign runs on impressions from publishers across the web, on every device.

The breadth and diversity of this inventory can be quite intimidating. How do you make sense of this fragmentation, take control of your buy, and leverage PMB to your best advantage? The answer: through your audience.

Finding the right audience is at the heart of PMB; it’s the element that unifies the buy. When programmatic is executed well, it enables you to precisely reach only your desired audience, minimizing wasted spend. You’ve worked hard and allocated enormous resources to figure out who your consumers are. Buying programmatically simply empowers you to engage them — and only them — with your message.

The Evolution of Audience Targeting

To understand how targeting your audience across content can help drive conversions, let’s go back to its roots. The concept of audience targeting comes from direct marketing, where consumers’ relevant characteristics (like automobile registry data on who owns a luxury car, or aggregated credit information about who is a good risk for a credit card) were linked with their contact information (like postal addresses) to form a database of consumers more likely to be interested in a brand’s products.

The result may have been annoying — junk mail in the mailbox — but it was delivered thoughtfully, with metrics like “open rate” that allowed marketers to use the channel effectively and efficiently.

Around 2008, marketers began incorporating audience targeting — from offline and online sources — into banner ads.

How did they do it? Data companies created match networks that linked their consumer databases to browser cookies. The match networks consisted of web publishers who, for a fee, anonymously linked data attributes such as “owns a luxury car” to browser cookies.

In 2010, after the Interactive Advertising Bureau established video standards for the digital ecosystem, these data-targeting assets could also be deployed for digital video, or broadband, advertising. Extending these capabilities into TV and mobile was more difficult.

In the years between 2008 and 2013, several cable and satellite companies adopted advertising technology that enabled TV targeting at a household level. While the process and technologies have matured, the reach of these technologies is still limited to some 20 million households, a fraction of the 116 million TV households in the U.S.

Mobile was a notoriously difficult problem for marketers focused on audience. The scramble of web and app ad inventory was further complicated by the different standards of the Apple, Android, and other operating systems. Only since (roughly) 2012 has a targeting technique known as “statistical IDs” allowed marketers to extend their broadband audience targeting capabilities into the mobile ecosystem.

In 2014, marketers now enjoy a wide scope of opportunities for audience targeting across online display and video, television, and mobile. Virtually all these screens and formats are accessible to PMB.

Sources of Targeting Data

The types of information available for targeting are as nuanced as humans themselves. Information can come from several sources:

  • CRM data, sourced from a brand’s own databases of offline purchases or e-commerce shopping and browsing.
  • Past purchase data, aggregated across sales channels by third-party companies.
  • In-market data that either models past purchase behavior (“it’s time for household XYZ to buy another sedan”) or collects signals from digital shopping and product research that indicates an impending purchase.
  • Rich demographic data is also available both from offline and online data companies. This information covers more than age and sex, to include factors like ethnicity, income, credit-worthiness, education, etc.

These data sources combine to paint a rich portrait of an audience who, if they had been targeted through multiple publisher buys, might be prohibitively expensive to reach. Yet, purchased programmatically, a brand’s chosen audience may be aggregated into a single buy, and made accessible through technology as if they were the audience for a single publisher or network.

Challenges of Audience Buying

While this process has evolved greatly in the past six years, there are still challenges. First, the data you use is only as strong as its source, and virtually any source will have flaws: data may be stale or outdated; it may be modeled and hence contain a margin of error; and through the many handoffs between the data source, match partner, and media platform, it may be matched improperly.

Second, data may lack scale, which will inhibit reach. Database matches are prone to match rates of 30 percent on the low side and 75 percent on the very high side, so that erosion will take place.

Managing Audience Data

Brands should employ certain strategies to manage the challenges and capture the opportunities of audience buying through programmatic:

  • Validation. To manage the risk of inaccuracy, brands should test their data providers using a market research company or panel, and compare the results of different providers.
  • Modeling for Scale. Brands should choose technology partners who have the capability to build machine-learning models to expand the reach of the data they use outside the footprint of the standard match partners.
  • Consistency Across Screens. Now that an element of consistency is possible across television, desktop, and mobile, brands should seek out partners who can provide the reach and coordination afforded by new technology.

Employing these tactics, your PMB should enable your ad to run in diverse types of content, platforms, and devices, so you can find the luxury auto owner while he’s researching his next car on his tablet, the stay-at-home mom who is searching for the perfect vacation destination on her laptop, and the millennial looking at different lipsticks on on her iPhone. This is how you programmatically reach the right audience at the right moment — anywhere and everywhere they are.

Justin Evans is the chief strategy officer at Collective. Email him at


"Programmatic Fuels the Audience Engine." Justin Evans, Chief Strategy Officer, Collective. ANA Magazine Spotlight. August 2014.

You must be logged in to submit a comment.