Contextual Targeting’s Coming Renaissance

5 min readMar 14, 2018


As Behavioral Tracking Comes Under Pressure, Look Forward to a Revolution in Contextual Targeting

Myles Younger, Director of Marketing, MightyHive

Cookies under siege

The recent anxiety around Safari’s Intelligent Tracking Prevention (ITP) and the EU’s General Data Protection Regulation (GDPR) is only the latest chapter in ad tech’s long-running epic tragedy: “The Death of the Cookie.” Let us commence the collective wringing of hands. A few years ago we could write off a handful of ad-blocking misanthropes, but now entire continents have turned against us! No one’s going to opt in! Without user tracking, how will programmatic advertising even work? We’ll lose all the amazing progress we’ve made! Won’t someone please think of the pixels?

“The Data Processors are holed up in that castle! Get ‘em!”

Specifically, what’s coming under pressure is behavioral targeting, which works by tracking user behavior on browsers and devices. But frankly, the Behavioral Tracking Arms Race was getting a bit myopic. So we won’t be able to carpet-bomb the internet with commodified octo-box dynamic retargeting ads any more. B.F.D. Are we quite sure that was the apex of digital advertising?

When the ad tech gods close a door, they open a window

Meanwhile, contextual targeting is ripe for a renaissance. While behavioral targeting has been hogging all the glory, contextual targeting hasn’t changed that much — you buy fixed topical or thematic placements that reach a rough approximation of the people you think will buy your stuff. Not much different than TV, print, radio, and outdoor. A bit old-fashioned, no?

This simplistic approach utterly misses the full scope of “context” in the programmatic age. Programmatic context doesn’t have to map to one-dimensional classifications like “fashion” or “movies.” Context is literally every quality and attribute of both the digital world (e.g., the content being consumed or the device type) and the real world (e.g., location or time of day) woven together at a moment in time when a person sees an ad.

Perhaps a cherry-picked historical parallel will convince you that we’ve barely scratched the surface of programmatic contextual targeting…

Back when Yahoo! was the bee’s knees circa 1990-something. So cute!

Remember when Yahoo! used to manually assign every web page to a hierarchical category? Their Dewey Decimal System for the web was the state of the art at one time (to be totally, overly clear: I am equating mid-90's Yahoo! to present-day contextual capabilities).

Then Google came along, pioneering a fundamentally new responsive, scalable, intelligent, and dynamic approach to search. Yahoo! was obsolete overnight and Google has been happily minting cash ever since. Likewise, the next generation of contextual targeting will be a radical departure from the status quo.

TV and why context matters

So who gets the conversion credit, the TV or the newspaper?

Let’s travel further back in history to examine the power of context. Why does TV remain the most powerful advertising medium ever conceived?

One reason might be that the context in which a TV ad was viewed (before the internet) was incredibly consistent. Someone was usually in their home, generally relaxed, and watching entertainment via a linear medium. There was no ad-skipping or binge-watching, and shows aired at specific times of day.

Culture evolved on a national scale to accommodate TV. Hundreds of millions of people shared a single context for hours and hours, day after day, year after year (otherwise known as “primetime”). It was a mass marketer’s dream. With a low variance in context, the Mad Men could laser-focus on understanding which products worked best for TV advertising and what types of people watched which shows. They fine-tuned the messaging and set advertisers’ cash registers ringing.

The infinitely fragmented context of digital

These people look REALLY receptive to advertising.

Fast-foward to present-day digital advertising. Do you have very much understanding of your ad viewers’ general disposition? No. Your DSP doesn’t really know what people are doing or how they’re feeling.

The Mad Men of old might chuckle derisively at your blurry view of context. You might have a billion cookie IDs with user-level data across hundreds of dimensions, but all that tracking data gives you close to zero understanding of a person’s world at the instant they consume your ad (or whether it’s an opportune moment to deliver an ad at all ).

It’s safe to say you’re probably widely missing the mark on context with most of your digital media dollars, and when you do happen to get it right, it’s just luck.

What does next-gen contextual targeting look like?

Looking to the future, contextual targeting has the potential to become more powerful than behavioral targeting ever was. But we first have to expand our definition of “context.”

Contextual targeting has heretofore focused mostly on simplistic objective attributes (e.g., “travel”) that are easily understood by computers. However, advances in artificial intelligence now allow software to process and understand the subjective (moods, emotions, attitudes, aesthetics — the elements of human psychology that form the basis for successful advertising). Software that can make sense of nuanced “shades of grey” will give marketers completely new tools and territory to explore.

If there’s one thing holding your campaign back, it’s the inability to target against Approbativeness.

Machines with the ability to interpret the subjective contexts will revolutionize contextual targeting. Great brands are built by selectively attaching themselves to the subjective — to experiences, symbols, and emotions. Are you consuming content that’s upbeat or dreary? Simple or mysterious? Cosmopolitan or crude?

But contextual signals aren’t limited to the digital realm. Every time a person is exposed to an ad, they are inhabiting their own personal real-world context — a context that only lasts for a moment in time. As you read this, the world around you is defined by weather, location, time of day, geography, local demographics, local and world events, even other people nearby. The ability to target simple real-world contexts such as geolocation is certainly possible today, but there’s still plenty of room for innovation.

Gradually, we’ll unlock contextual signals across the digital world and the real world, across the objective and the subjective. And as we do so, we’ll open up incredibly rich moment-in-time opportunities to serve the right messages to the right people at the right time. And if I’m not mistaken, that was always the promise of all those cookies.