Home Online Advertising The Crawl, Walk, Run Guide To Contextual Targeting

The Crawl, Walk, Run Guide To Contextual Targeting


Comic: Dusting off a classic.Contextual targeting online dates back more than a decade. But the technique is having a renaissance as marketers test out strategies that work in a cookie-free world.

In contextual targeting, a machine scans page content and places a digital ad relevant to that context. Natural language processing allows contextual tech providers to understand nuances of an article, including figuring out homonyms (Apple and apple) and determining sentiment. 

Then the content must first be categorized by a contextual targeting vendor. A few different flavors of companies exist. Brand safety and verification companies, like DoubleVerify, Integral Ad Science and Oracle Contextual Intelligence (formerly Grapeshot) offer contextual targeting. There are also standalone contextual companies, like GumGum, Peer39 and Zefr. Media.net operates a contextual-focused SSP. And Major DSPs, like Google and Amazon’s, also offer built-in contextual targeting. 

These companies may use the IAB’s standard content taxonomy, which covers more than 700 types of content – from budget travel to bowling to gospel music to baby showers. Other tech companies offer dynamic categories based on AI or even more granular keywords – which enable buying not simply articles about smartphones, but ones about Apple.

Contextual in action 

Applications of contextual targeting range from obvious to clever.

Context can make a creative message pop. For instance, the UK-based paint company B&Q Valspar showed ads for colors that matched the colors of images in the article with contextual targeting company GumGum’s tech.

Context can also signal intent. An appliance brand targeted content that people tended to read as they stepped toward buying its products, and saw 50% gains in performance compared to its normal targeting parameters because it was able to “hone in on an area of the conversion funnel,” said DoubleVerify Chief Product Officer Jack Smith.

In this application, context is trying to accomplish the same thing as behavioral advertising – just with a single data point.

Contextual targeting also helps marketers in highly regulated industries show up in the right place in the right way. “There are certain categories, like financial institutions, pharma and alcohol, where you are going to see more contextual because of limitations from a regulatory standpoint,” said Ryan Eusanio, director of programmatic marketing operations at Omnicom Media Group. 

Finally, first-party data can be added to context. Using dynamic creative optimization (DCO), a company can talk to their customers in the right context – with a message tailored for that context.


AdExchanger Daily

Get our editors’ roundup delivered to your inbox every weekday.

Crawl: Set it up!

At the crawl stage, contextual advertising is usually a line item on a media plan – an isolated tactic. 

Brands have likely already been doing some form of contextual targeting. Brand safety, for example, uses the same underlying tech as contextual advertising. The non-techie version of contextual – doing direct-to-publisher buys like ESPN’s football coverage or a news publisher’s sports section – is usually an existing component of a marketer’s media plan.

Buyers who want to speed up their crawl in contextual should focus on contextual categories related to their brands, said Erica Sklar, EVP of digital investment at Zenith.

So an airline should advertise against a travel article instead of one about green energy, where they’d need to consider a different message.

The buyers should also set up the campaign in a way that gives them a leg up on future measurement and optimization.

Even when starting with broad categories, buyers should try to slice up their buy so they can see how certain keywords or types of content – like the Super Bowl, fantasy sports, individual players or teams – drive results.

“We try to be as granular as possible when we set up campaigns, so we can see how each subcategory performed,” said Prerna Talreja, group director of marketplace for independent media agency Crossmedia. A yoga brand targeting a fitness context might get better results positioning their messaging against meditation articles, for instance, rather than topics around high-intensity exercise. Testing multiple categories allows for better optimization.

However, adding contextual categories can potentially add cost and reduce scale. Brands need to think about these details as they test.

“Decisions can’t be made in a vacuum of price, cost and scale,” Eusanio said. “You have to understand what kind of impact it will have, preferably ahead of time.”

Walk: Test, rinse, repeat

In the walk phase marketers and agencies can bring more rigor to how they measure success of their contextual campaigns. They can start optimizing the contextual categories where they show up, and see if they can squeeze out better performance from a different vendor.

One big challenge of testing vendors head-to-head in the same categories is that every provider can group content however they see fit. What counts as “beauty” or “sports” content varies by tech company, which, in turn, can impact the accuracy of the segments, their scale and their performance.

“The algorithms that these companies have built are sophisticated, but they’re also not 100% transparent,” said Crossmedia’s Talreja. “Classification is very subjective.”

Give the same contextual targeting company one hundred URLs, and each one’s algorithms will likely classify the content of the pages differently. Buyers can test tech by asking contextual companies to perform this exercise as a way of testing this tech, said Derek Wise, Chief Product Officer at Oracle Data Cloud.

But accurate categorization may not correlate with performance. The most sophisticated contextual provider may not deliver as strong performance as a “good enough” contextual solution, Sklar said (with the caveat that the tech can’t be so wrong as to cause brand safety issues).

“At the end of the day, what drives the best KPI counts, even if a provider returned a beauty page as a celebrity page,” she said – with the caveat that the context must still meet brand safety standards.

Run: Context and DCO

Contextual targeting starts to get really interesting when brands use the best-performing context to inform creativeContextual advertising messaging.

“DCO [dynamic creative optimization] has a really strong fit with contextual, especially if we lose some of the other signals we can put in our decision tree, like audience targeting,” Zenith’s Sklar said. “Context opens up an opportunity to get more personalized through the environment, not the audience.”

Brands can create multiple versions of advertising creatives for different contexts, and in turn serve those ads to different sets of their customers – a “muscle memory” most agencies already have from using the same technique with audience data, said Max Jaffe, global head of programmatic at GroupM. “Keyword categories can inform the creative messaging and better tailor the message to the environment they’re in.”

In this phase, context isn’t just a way to tailor a message and optimize where it runs. Brands should use the results to generate broader insights that inform overall strategy. Success with certain subtopics could inspire new creative messages and give brands insights about their customer journey.

In the run phase, marketers are also thinking deeply about where contextual could complement – or replace – audience targeting. Brands may test audience targeting with contextual targeting side by side, or layer them on to drive better performance. With a year before cookies go away, now is a good time to figure out what context –if any – could serve as a solid proxy for third-party audience data. 

What’s ahead in contextual targeting

While brands are leaning into contextual targeting in 2021, the technique is still playing second fiddle to another alternative to audience targeting: a first-party data strategy. 

“Brands are going to need a cocktail of deterministic, probabilistic and contextual targeting,” said Matt Prohaska, CEO and principal of Prohaska Consulting. “ 

Many brands – rightly so – are prioritizing developing data about their customers that’s uniquely valuable to them. But contextual targeting can also play nicely with a finely tuned first-party data strategy. Contextual targeting is also a cheaper alternative to audience data, and that lower cost could help brands.

One big challenge that stands in the way of contextual targeting remains: measurement. Though the technique doesn’t invade users’ privacy, the end of third-party cookies means it will be impossible to measure the impact of contextual advertising on conversions, a Catch-22, Talreja said.

Since measurement drives media investment, murkiness in attribution could keep brands from expanding their use of contextual advertising. “It’s a blind spot,” acknowledges Peer39 Chief Operating Officer Alex White. “Even though it’s the ideal environment, you still have the gap in measurement.”

Must Read

It’s Open Season On SaaS As Brands Confront Their Own Subscription Fatigue

For CFOs and CEOs, we’ve entered a kind of open hunting season on martech SaaS.

Brian Lesser Is The New Global CEO Of GroupM

If you were wondering whether Brian Lesser was planning to take some time off after handing the CEO reins of InfoSum to Lauren Wetzel last week – here’s your answer.

Comic: S.P. O'Middleman's

TripleLift CEO Dave Clark Abruptly Exits After Setting The SSP On A New Trajectory

Dave Clark, who’s led TripleLift for the past two years, is stepping down, effective immediately, and is being replaced by a coterie of TripleLifters.

Privacy! Commerce! Connected TV! Read all about it. Subscribe to AdExchanger Newsletters
shopping cart

Moloco Invests In Its Competitor Topsort As The Retail Media Stakes Go Up

Topsort can lean into Moloco’s algorithmic personalization, while Moloco benefits from Topsort’s footprint with local retailers in the US and in Latin America.

CDP BlueConic Acquires First-Party Data Collection Startup Jebbit

On Wednesday, customer data platform BlueConic bought Jebbit, which creates quizzes, surveys and other interactive online plugs for collecting data from customers.

Comic: The Showdown (Google vs. DOJ)

The DOJ’s Witness List For The Google Antitrust Trial Is A Who’s Who Of Advertising

The DOJ published the witness list for its upcoming antitrust trial against Google, and it reads like the online advertising industry’s answer to the Social Register.