Contextual targeting has historically involved buying impressions on pages whose content is deemed relevant to a marketer’s brand or product, using a process called keyword targeting.
For years it’s been seen as a safe bet – advertisers can control the content their brand appears against and increase the likelihood of reaching users with relevant interests. For example, an airline might want to appear against content around city breaks or family holidays.
But the traditional points of ease around contextual have also held it back – it’s been seen as probabilistic, limited in scale and cluttered with brand competitors.
Thanks largely to recent developments in AI and machine learning, these shortcomings are now being dismantled one by one, and as third-party cookies decline in usefulness and viability, modern contextual is gaining momentum as a possible alternative.
So, what do you need to know?
From keywords to key moments
While contextual targeting usually works without relying on cookies or personal data, concerns highlighted by the ICO last year around sensitive data suggest that certain keyword targeting may in fact pose a compliance risk. This is still being investigated, but it’s due to the possibility of users being targeted (or not targeted) based on personal assumptions made about them from the content they are consuming.
As a result, advertisers are increasingly looking beyond keywords to more sophisticated and nuanced methods of understanding context and when it might be influential to a campaign. Engaging with users in the correct contextual ‘moment’ is beginning to emerge as more important, effective and compliant than targeting them due to the specific topic they are consuming.
Free from the limits of keywords, modern contextual methods are now able to deliver scale by targeting pages across a range of brand-safe topics – expanding to find new audiences, rather than narrowing. This agile deterministic scaling removes the guesswork from contextual, as the decision to place an ad on a page is triggered by live brand engagements or attention clusters which are generated by a brand’s consented first-party users. Driven by engagements and attention metrics in this way, rather than by topic keywords, a campaign can explore an almost endless number of quality online pages and is unlikely to run out of road.
Standing out from the crowd
Keyword targeting can leave brands vying with their competitors for impressions which don’t necessarily deliver results. Our work at Illuma shows that as many as 93% of brand engagements take place away from obvious contextual environments, where of course the chances of brands running alongside competitors are greatly reduced. This finding also aligns with the fact that consumers are multi-dimensional and sophisticated individuals with varied interests, who might engage with an ad in any number of quality contexts, as long as it is served in the right online moment for them.
Audience and the power of first-party data
As our understanding of audience behaviour evolves, it seems natural that the technology should follow. Advances in AI are allowing advertisers to analyse and respond to the real-time content consumption of a brand’s first-party users and in the process, unveil information about the content which is driving ad awareness or engagement.
These audience insights are often unpredictable and challenge not only where in their browsing journey a potential customer might be, but who they might be. This information almost always offers advertisers new, complementary and surprising information which might go on to inform other areas of their marketing activity.
Blocking the ad blockers
Developments in modern contextual targeting are also benefiting publishers by slowing down overzealous ad blocking which, for example, have prevented ads running against words such as Sussex (it contains the word ‘sex’), ‘shoot’ (could refer to a film or a weapon) or ‘hash’ (the breakfast item or the drug). Mass ad-blocking is ruling out the monetisation of vast swathes of editorial collateral for publishers and potentially placing their news operations in financial jeopardy.
Advanced contextual algorithms are now able to perform deep-reads of editorial pages and give publishers truer, more nuanced understandings of the full range of contextual signals coming back from a page and the moments that work, in real-time, for their advertisers. In due course, publishers should also enjoy increased revenues by working with this sort of technology, as these deeper contextual algorithms naturally favour quality, long-form content.
To find out more about advanced contextual targeting, contact us.