In recent years, there has been a lot of talk about the growth of “contextual targeting”, especially in the online space. Contextual targeting focuses on the context in which users see an ad, compared to audience (or behavioural) targeting which focuses on ensuring the correct user sees the ad, regardless of context. Whilst there isn’t a conclusive answer as to which strategy is more effective, new regulation such as the EU’s GDPR policy and the slow demise of the once-celebrated “cookie” has many in the advertising industry believing that context will be key come the next few years, as complex compliance issues arise around gathering and using user data.
From what we understand about contextual targeting, using this strategy generally benefits advertisers and campaign effectiveness. One WARC study found that “media-context factors seem to elicit higher recall and recognition for advertising”, meaning users are more likely to remember ads if they are served in a contextually relevant environment. GumGum Inc. – an AI company specialising in solutions for advertising and media – also carried out neuro-analytic research examining how an ad’s contextual relevance to content influences consumers. Their findings showed that ads placed in the most contextually relevant pages drew 43% more neural engagement and 2.2 times better ad recall, supporting findings made by WARC and many other researchers in this field.
It’s clear that context matters, however there are still some shortcomings in the industry when it comes to utilising the full power of context, with these limitations posing problems for publishers and advertisers. Just a few weeks back, The Guardian reported that UK newspapers could lose up to £50m in digital ad revenue as advertisers blacklist news sites due to the large amount of coronavirus related content. Advertisers also blacklist certain keywords, meaning inventory is going unsold even in contexts that are perfectly brand safe. The Coronavirus pandemic has meant that publishers are seeing record levels of online traffic, but programmatic pricing is down due to lower CPM’s and CPV’s. Advertisers are worried about aligning ads relating to coronavirus content, meaning inventory is going unsold, causing revenue problems for publishers and media agencies, whilst advertisers are missing out on reaching large, engaged audiences.
This pandemic has only highlighted a problem contextual targeting has faced for a while and one which is starting to be addressed. Advertisers at times are unnecessarily missing out on quality inventory because the contextual methods currently in place are lagging behind. We know through research that negative content can hurt ad recall, but we also need to use ad-technology to run smarter contextual campaigns if we are to assure advertisers that campaigns will perform well. This can be done in several ways, but below are three factors I think will be of importance as the industry moves forward.
Move beyond text
Firstly, I believe it is important to utilise contextual intelligence and AI technologies. Often keywords are used to determine whether a particular site or page is suitable, however by using more intelligent systems we’re able to take into account the whole context of the page including audio, images, video, and more. This will help campaigns distinguish what pages are actually “brand safe” and suitable for ads, benefitting all parties.
Advertisers/planners can sometimes blacklist entire categories or ignore certain environments when planning campaigns. Although prudent, this can mean campaigns miss out on reaching appropriate, qualified users. News sites are a great example as they often host content which is not considered “brand safe”, even though many pages/sections are completely appropriate for ad placement. By using more intelligent contextual methodologies, planners can refine contextual targeting strategies during the planning phase so advertisers do not miss out on quality impressions.
Don’t ignore the audience
This is a crucial point as your audience always matters, and utilising audience data in hand with contextual strategies ensures less wastage as relevant users are reached. Audiences are complex, and contextual targeting alone is not always the best strategy as a lot depends on the product, brand, and ad itself when measuring ad recall. The WARC study I mentioned earlier also found that contextual ads provide greater ad-recall for the ad/product itself compared to the brand. Dependent on your KPI’s, this can be both a good or a bad thing.
Contextual targeting is difficult to perfect, but the industry is only going to get better at reaching the best audiences. As always, multiple factors determine whether campaigns perform well (context is just one element), but the better we can refine the planning process for contextual targeting, the more the ad industry will benefit.