By Luc Benyon, Head of Marketing, Video Intelligence
“The brain can choose what it pays attention to,” Neuroscientist Dr. Ali Jennings tells us, “if you are reading a piece of content and someone has put an advert in there that has no relevance to you, you will simply not even see it. Your brain does not allow that information in.”
It’s a story all too familiar to online users, and it can’t continue. In 2019, putting the context back into advertising is one of the first steps to the digital ad world looking to rescue its reputation.
Here, we’ll use examples from social media, video formats and programmatic advertising to provide a glimpse into the melee of digital formats that media buyers are faced with. Each has their own nuances, and each offers opportunities for advertising to operate contextually. Let’s start with the format that probably led you to this article: social.
On social media, context is dynamic; the experience of seeing a social post will differ for every user. There are a multitude of variables affecting how a post appears and is perceived. For instance, the other posts in the feed frame a post. They are determined by an algorithm, preference, and previous engagements. Similarly, likes, shares and comments also each provide different elements of context. So social media a huge challenge when it comes to context.
For anthropologist Alexandra Georgakopoulou-Nunes, this is understood as ‘context collapse’. Whereas in previous generations we shared stories socially among tight groups, organised around work, family or school, now these different contexts have collapsed together online.
For advertisers, ‘context collapse’ encapsulates the risks that are entered into whenever social media is part of a campaign. Knowing who will see your post does not help know how it will be seen. What other posts surround it, comments, and other engagements will all change the perception of it. These are unwieldy, uncontrollable factors that make social media the nexus of collapsed context.
On the web, format and planning are the biggest factors when it comes to context. Too often digital advertising plays fast and loose with context.
Programmatic buying has been responsible for some of the past few years most mis-placed media, as buyers blindly buy inventory across the open web, leading to potentially awkward placements. It’s the main cause of brand safety issues — frequently listed as one of marketers top concerns. As Dr. Jennings and neuroscience shows, this experience is a misnomer for our brains. As context is lost, so too is much chance of engagement from the user.
Meanwhile re-targeting banners bring ads for previously browsed products and serve them completely out-of-context. We’ve all experienced the bizarre ad for a holiday home appear on the side of a page about global politics and thought, ‘WTF?’. Retargeting is perhaps the clearest example of disregarding context in favour of user-targeting.
Thankfully, to most web user’s relief, GDPR has made retargeting and programmatic less effective. New methods of understanding and reaching users are emerging.
One of the joys of digital marketing is its constant adaption to market forces, and relentless push to innovate. Contextual advertising is currently benefitting from this.
Native advertising encompasses formats where the ad unit feels more like it ‘belongs’ to a page. For that reason it is rightfully popular, but we need to make sure the content is relevant too; too often at the moment, native units are filled with irrelevant clickbait.
Media buying techniques like Private Market Places [PMPs] allow media buyers to group sites together in categories. This goes some way towards targeting their ads onto sites where the content is relevant to the product. It also reduces the randomness of the ‘open market’ whilst retaining the speed and scale media buyers demand. PMPs will grow in prominence as advertisers look to control their buys more.
Machine-learning-based techniques such as natural language processing, semantic analysis and distributed search are revolutionising the way we can understand page content. As a result, new products have emerged which apply machine learning to contextual advertising. In practical terms, this means it’s now possible for content and advertising to be placed within pages, based on ‘context’, rather than targeting user behaviour.
So, as we begin to understand context more, we’ll see more emerging formats, buying techniques and technologies that deliver advertising contextually. We’ll retain the advances in speed and scale, thanks to automation, and see better engagement with our ads, and happier users.
Join Dr. Jennings and Professor Georgakopoulou-Nunes and video intelligence on March 20 at Advertising Week Europe, where we’ll reveal the science behind some of these claims in greater detail.