Context, sentiment & attention in targeting’s present & future

The rise of contextual targeting has been percolating as a trend since the arrival of GDPR in 2018. Arguably, the major tech giants’ ongoing response to legislation, plus public opinion, has propelled it even further into the spotlight.

From Apple enforcing AppTrackingTransparency on mobile, to major browsers limiting cookie usage both in the present and future, non-ID based targeting methods would appear to be in the ascendancy. Perhaps it’s not a great surprise that 48.7% of advertisers worldwide are now saying “most of our buys” in 2022 would rely on contextual targeting, according to DoubleVerify research.
 

From keywords to sentiment

Another important consideration is the evolution of the tech driving contextual over the past couple of years. 

A short time ago, context was little more sophisticated than parsing and categorising page keywords. But now more advanced technologies - like our own here at Nano Interactive - can understand not just the underlying meaning of content, but also the sentiment of a given article. And this extends to every person, place or product mentioned within content, as well as the piece as a whole. By extension, with the added benefit of programmatic, we can then interpret the evolving intent landscape and topic sentiment over time.

This brings into range an overall more nuanced approach to context than became widespread in the keyword targeting era. For example, optimising mid-flight to negative sentiment content across competitor targeting, if that is driving stronger results. It also means we can move towards a smarter approach to topics - not ruling out entire content categories or keywords - leading to serious issues in scalability, especially when targeting is tight. Instead, we simply prioritise more by sentiment.

Rule-based sentiment targeting also allows brands and agencies to protect brand activity with positive and neutral sentiment rule overlays, layer sentiment targeting, and conquest consumers reading content that mentions competitors and competitor products with negative sentiment. Meanwhile, you can build positive brand association across broad topics by layering on positive and neutral sentiment targeting.
 

From sentiment to emotion

While we will never know for sure consumers’ emotional state, the continued evolution of contextual targeting is likely to be in this area. Tracking emotion within content offers marketers an even deeper understanding of their user by looking at how the emotion behind content searched for and consumed changes over time.

As with sentiment, emotion is measurable both at an article and topic or entity level. And when used alongside sentiment, it can be tied to further insights around brands and topics. For example, let’s say we can see that users are searching for and consuming more content on vegan beauty products. How does this help brands start to plan campaigns, or inform creative messaging for advertising campaigns?

From CTR to attention metrics

Content as a proxy for intent is a live, observable fact at scale. While the cookie faces challenges around coverage, and may be indicative of intent up to a month old.

Rightly or wrongly, cookie-led activation has arguably also led to an overwhelming focus on clicks as a measurement and optimisation tool. But the shift to advanced contextual techniques outlined above is also presenting new options and ideas here - for one, a re-evaluation of attention metrics.

We commissioned Cint and Sapio to do just this, comparing click-through rate (CTR) versus attention optimisation within the same campaign for a gambling advertiser. The results suggested that attention might significantly improve ROI, especially around branding: a 121% comparative increase in consideration, plus a 58% greater brand perception score were just two of the results.

In summary, from GDPR in 2018, to Apple’s ATT last year, and the cookie changes affected by Safari and Firefox already, and Chrome’s to come, the direction of traffic seems pretty clear. Changes at an industry level are of course challenging, but if there is a positive, they are also forcing us to change and re-evaluate how we work. One part of that shift may just be towards real-time intent, in the here and now – with context, attention and sentiment all with a role to play in targeting’s future.