In recent months it’s been said that context trumps audience for content promotion and native ad targeting. Some might have you believe that the “where” of native targeting is more important than the “who.” While certainly an interesting point of view, I’m not so sure I completely agree.
At some level context is certainly an entrée to a well targeted audience. But if you can successfully target an audience that’s worth engaging, then why wouldn’t you want to extend and repeat that engagement with the same audience? As it turns out, audience based targeting is a great way to do exactly that. With the latest audience targeting and retargeting capabilities, it’s now possible to sequence content – delivering repeat exposure for a variety of content, based on data and signaled consumer preferences.
Now admittedly I have a bit of a strong opinion when it comes to contextual targeting. As an 8-year Googler, I launched Google’s AdSense for Content business in 2003, built a 20-person product management team and grew it to a multi-billion dollar product line. Context sells, and context serves an irrefutably valuable purpose. But here’s why I’m more excited about the combination of context and audience based targeting versus context alone:
- Specialty properties (e.g. trade publications like the one you’re reading) are only a small fraction of total pageviews on the Web and are typically very expensive placements as they provide “the right user, at the right time” context. These are highly valuable placements especially for lower funnel campaigns, but frequently too scarce or expensive (or both) for effective, long-lived native exposure at scale.
- Category properties (e.g. tech news, auto news) are a much bigger fraction of pageviews on the Web and are more in line with the pricing of a content promotion campaign. On these pages, a blend of context and audience targeting are best. Audience works fine here since users are frequently in discover and browse mode versus research mode.
- General interest properties, aka “the contextual desert” (e.g. general news, sports, social media, etc) make up the vast majority of pageviews on the Web. In this contextual desert the best targeting signal is what you already know about users’ interests. Audience targeting results in a better ad experience than attempting contextual targeting for this large swath of inventory.
To put all this in perspective with a real example, consider a retail food brand we’ve been working with this past year. They have a staggering portfolio of recipes, but realized they needed to flip their content-audience dynamic from a users-per-recipe orientation to a recipes-per-user model, without sacrificing scale. By starting with context and following through with audience based targeting and retargeting, they were able to increase repeat engagement by 4x. Repeat users read an average of more than 5 recipes! Trying to deliver that level of repeat engagement by waiting for repeat users to visit contextually related pages (to show them interesting content) does not scale and would be a disservice to users and advertisers. Advertisers would miss most of the opportunity to reach the user, and users would see advertising that aligns with the context of the page but not their more strongly signaled interests.
So let’s use contextual targeting for its purpose and ability: a well understood way to discover users’ interests. But when it comes to content promotion and native advertising, let’s also think beyond raw “distribution” and start to build actual relationships with individuals through new models for audience based targeting – which in my biased opinion is one of the best and most exciting capabilities afforded by today’s global ad tech ecosystem.