The Myths (and Truths) of Contextual Targeting

Contextual targeting is enjoying a resurgence this year, but unfortunately, we've found that along with big headlines, a few myths are still making the rounds. Here are some of the untruths we've decided to clear up.

Myth #1: Contextual targeting is just keywords

While early contextual targeting providers relied on using keywords to find friendly territory for advertisers, the technology has evolved far beyond those nascent roots. In fact, we would go so far as to say contextual targeting providers who still rely only on keywords are doing their clients a disservice. If, for example, you're managing advertising for a cooking brand that only wants to run alongside comforting food content, your keyword anti-targeting list would flag the word ‚Äúkiller,‚ÄĚ costing you a perfect placement alongside a recipe for ‚Äúkiller key-lime pie.‚ÄĚ

When evaluating partners, if their technology has evolved to include the wide variety of artificial intelligence technologies that would allow for page-level analysis of body copy, pixel-by-pixel image analysis, and sentiment analysis. Incorporating technologies like natural language processing and computer vision‚ÄĒas GumGum does‚ÄĒ should be table stakes by now.

Myth #2: Contextual targeting is a blunt tool

Perhaps a consequence of our first myth, too many people think of contextual as a blunt force tool. If a brand wants to run ads for high-tops, they target basketball-related content. If they want to avoid politics, they program their campaign to avoid the names of key politicians.

But as we outlined in a previous column, advertisers can use contextual advertising to run strategic campaigns that target niche audiences or lookalike audiences. Consider our work for a major streaming service debuting a revival of a young adult drama. A blunt-force approach might be to simply run ads against the title as a keyword or IAB contextual categories that appeal to teens. Using the GumGum platform, however, the team targeted not just content with positive mentions of the client's new show, but also conquested content about shows with a similar target audience and content that might appeal to older audiences looking for a hit of nostalgia. With a sharp strategy team in place, contextual proved to be a cutting-edge tool.

Graphic showing how keyword contextual targeting is old fashioned

Myth #3: Contextual targeting is more expensive and less effective

Marketers have such a rich understanding of their target customers‚ÄĒinterests, passion points, motivations, and more‚ÄĒthat they hardly need to reduce them to audience data points. And yet the narrative that audience targeting better meets marketers' needs stubbornly persists.

Studies show that contextual marketing is both more cost-effective and more engaging than behavioral targeting. In a recent study with the Dentsu network of agencies, we saw that category-wide, contextual advertising saved advertisers 29% on a CPM basis, and GumGum's Verity platform delivered those CPMs for 36% less than behaviorally targeting. We also saw an uptick in relevance that yielded a 48% savings in cost-per-click when Verity served ads.

Myth #4: All contextual is created equal.

Not all contextual vendors are equally matched. To start, not all incorporate the same kind of artificial intelligence outlined above. The proof is in the data. We found that 67% percent of content blocked based on COVID-19 brand safety concerns was perfectly safe. Advertisers missed out on millions of dollars in inventory because they relied on tools with sub-par technology.

Test potential partners' skillset with a spreadsheet of 1,000 web pages and incorporate content you know would be safe and suitable for your brand, but might be blocked by rudimentary, keyword-based solutions. Then have your partners determine which are brand-safe. Or, keep potential partners from grading their own homework by asking if they have been accredited by a third party, as GumGum has. This summer, the Media Rating Council verified that our technology is able to consider all available signals on a web page including text, image, audio and video, to determine if the context is safe and suitable for your ads.

Myth #5: We have time, let's max out on cookies

We know, Google has once again stayed the cookie's execution, this time postponing the phase-out until 2023. While some marketers might see this reprieve as an excuse to binge on cookies as long as they can, forward-thinkers should look at this as a chance to test and plan.

When Google drops the axe, it's sure to have a suite of opaque, walled-garden tools for marketers to play with. But like most tech titans, it won't be very transparent with data. By experimenting with contextual advertising now, marketers have a chance to build a portfolio of proven campaign strategies that they can use while tinkering with Google's new tools.


You might also like

Click this icon to scroll back to the top of the page