Contextual targeting has been used since long before the days of digital marketing arrived. In magazines and newspapers, marketers used content to determine context and choose where to place ads. As online advertising brought in new possibilities, we saw programmatic trading gain ground. Spend share jumped from 10.4 per cent in 2012 to above 70 per cent in 2021, according to Statista. Demand-side platforms exploited behavioural audiences, allowing for more streamlined and automated activation across thousands of publishers.
Third-party cookies strengthened the behavioural targeting proposition by pre-classifying users across multiple domains, making them available to be targeted across marketplaces. The different exposure prices for these users enabled the optimisation and reach opportunities that drive efficiency for brands and paid media marketers today. As a result, direct site buys became underused and contextual targeting receded into memory, used mostly for sponsorships, special activations or brands that sought greater control over brand safety.
But what we gained in efficiency and predictability we lost in user experience. Behavioural audiences are built based on how data aggregators decide to classify them. The normalisation of behavioural data signals by these aggregators contributes to the proliferation of a one-to-many experience rather than a many-to-one ad-serving model. This allows for more streamlined and automated ad serving but prevents more personalised and relevant experiences.
In the same way that online marketing brought advancements to offline media and opened up opportunities for businesses, behavioural targeting has done the same for contextual marketing. Networks and publishers that planned ahead for the end of third-party cookies, invested in building first-party data, migrated to newer technologies, and initiated the use of predictive analytics, are now able to deliver the same level of sophistication with contextual targeting that we saw with behavioural targeting. Data management and demand-side platforms have increased the number of events and properties being tracked, improving the analysis of context to deliver performance through better and more personalised user experiences. These developments will positively affect the user journey, as well as paid media campaigns, contributing to better engagement, attention and quality metrics.
This has already been reinforced through a recent study from Dentsu Aegis. They implemented a test to gain a robust understanding of potential best practices and tools available for success in a world with diminishing access to behavioural targeting. The study goes about testing different contextual intelligence vendors against behavioural targeting with two objectives: evaluate cost efficiency and compare accuracy. Verified by MOAT, Nielsen DAR, Xandr and Appen, the results were eye-opening, with contextual targeting producing 36 per cent lower CPMs and 48 per cent lower CPCs than behavioural targeting, reinforcing that contextual targeting is a viable alternative in a post-cookie world.
Furthermore, a study published by Dr. Erik Nylen, The Drum’s head of data science, reported results as staggering as Dentsu’s, establishing contextual targeting to be 4.7 times less expensive, deliver 12 per cent more attention, and yield significantly greater brand lift per second, being 7.5 times greater than the impact per dollar spent on audience behavioural targeting.
With a lower CPM and potentially higher performing engagement metrics, mixed in with diminishing third-party data points, contextual targeting and contextual intelligence platforms can not only outperform behavioural targeting, but also may become the new standard to deliver against performance objectives.
Brands and advertisers should start looking at performance differences between behavioural and contextual targeting to understand how their investment could be better used, as they prepare for a new era in digital marketing.
Start with determining what “good” performance means for your brand. Different techniques can be used to reach your objectives, but establishing benchmarks and understanding how your results could be better are vital to continuously improving performance.
Testing goes a long way. Whether it’s technology, people, processes or relationships, it can all affect the results your brand can achieve. Ultimately, performance is about correlations as much as understanding causations; map your approach out, understand the levers you can pull and how they will affect the type of performance you are after.
Finally, contextual targeting is more audience-centric and user-focused. It offers us as brands and advertisers the opportunity to deliver better experiences without sacrificing performance.
