By Carl White, CEO & co-founder at Nano Interactive
Ask yourself a question. When you’re looking for new customers, are you looking for people aged 35-44 with a central Birmingham postcode, who recently bought two pairs of trainers and a kettle and who visit more interior design websites than is strictly healthy?
Or are you looking for customers whose interests and politics largely chime with the media site they’re on, who share its dry sense of humour, compassion for humanity — and its obsession with the cottagecore aesthetic?
The former, in our privacy-first and impending cookie-less world, looks increasingly difficult to target. But then again, how much do you really want to? How much has this information told you about how to really grab this customer’s attention? You have personal data but little in the way of usable insight. Certainly, it gives you the ability to send them endless ads for trainers, which will stalk them round the web for months and drive the (now lost) customer completely potty.
The second example, however, lays the foundation for modern contextual targeting. By deeply understanding the nature of the media property you’re advertising on, you understand to a far greater degree what motivates or chimes with the people visiting it. That insight allows you to deliver ads that aren’t just targeted but ones that resonate much more deeply with the end-user, triggering brand awareness and purchase intent. Advances in content analysis and AI are enabling us to make major steps forward in matching medium and message to deliver advertising that works better than ever.
By matching the tone of the site and the individual page, advertisers are in step with their potential customer. This is significant as, according to Integral Ad Science (IAS) research, 72% of consumers feel their perception of an ad is influenced by the content surrounding it. If they are positively influenced, they are ‘people like us’ which increases attention, engagement and purchase intent.
Media and particularly mainstream titles, generally addresses broad swathes of the population. To truly understand how to use contextual targeting effectively, the advertiser has to delve into not just keywords and content but the sentiment of each piece, to understand the evolving intent landscape, find the right ad positioning and adjust the creative to match. A recent study by IAS found that ads designed to trigger an emotional response were 40% more memorable when placed on a contextually-relevant page. Being able to do this in real time, when the customer is actively considering a purchase, is a powerful tool.
Sentiment analysis uses natural language processing (NLP) to analyse the content’s degree of subjectivity. Looked at over time across a range of content, sentiment analysis can determine how an ad might be received if it were to be placed next to similar articles or blogs.
At a point where we have come to rely on the security that granular data provides, it can be hard to let go of the comfort blanket. In fact, many are refusing to, insisting that turning to tools such as Universal ID solutions will bridge the gap. These may well provide a bridging solution for now but they also add a layer of complexity to effective targeting that might be better avoided. Equipped with existing knowledge about customer segments, media properties and overlaid with sentiment analysis, marketers have the ability to target new customers as effectively as before – if not more so.
Sentiment analysis is a powerful contextual targeting tool but it’s made more so if advertisers are able to harness it to intent. By serving a relevant ad to a user in the moment they are expressing intent, they are most receptive to brand messaging. And by using artificial intelligence and machine learning it is possible to combine multiple live intent data signals such as how they’re navigating to a web page, their location or device. This insight is added to information about the environment the ad is served, including the content’s sentiment. Historical performance data about other ads in the same slot reinforces (or weakens) the argument for placing it.
So far, so effective but the additional critical element of layering sentiment analysis onto post-cookie targeting techniques is in engagement. When brands are competing for a wafer thin slice of consumer attention, being able to resonate emotionally is vital. Being able to talk to the customer with all the nuance of their own personality and values is key to capturing them in that moment. Being able to place ads that are attuned to sentiment delivers that emotional resonance.
The removal of cookies is a watershed moment for an industry that has spent the best part of the past two decades becoming hooked on heuristics for customer engagement that were often anything but. Sentiment based analysis and targeting brings back to advertising the understanding that human behaviour is often complex and contrary, meaning that delivering the right messages to them requires a subtle mix of art and science. An understanding that the right time and the right place needs to be matched with the right frame of mind.
