December 3, 2019
IIeX Behaviour UK 2019 was an interesting snapshot of how market researchers are adopting behavioral science. Many are aware of the cognitive biases affecting human decision-making and the need for behavioral measures. Here, we look at five conference takeout’s you need to know: what is behavior, and what is it not? The importance of understanding and measuring context, and finally what behavioral science can learn from copywriting.
Behavior can be understood as the way people act. However, when attempting to measure behavior, true behavior is rarely the focus. Stated behavior, aspirational behavior, emotions, attitudes or associations are not behavior, but proxies. Clients have an expectation around the term behavior, so, be clear about what you’re delivering. Don’t talk about ‘behavior’ unless you are measuring behavior.
Researchers are increasingly using emotion tracking as a proxy for behavior. The premise being that emotions drive behavior. Therefore, if you can measure emotion you can predict behavior. This is being used by the advertising industry to tweak creative work, so it elicits the desired emotion in the consumer viewing an advert.
What’s missing is any definite link between the target emotion and product purchase. Phil Barden discusses how emotional responses to adverts may lead to recall, reach and subsequent conversations, but the brand itself is no better remembered as it didn’t elicit the emotional response the advert did. Although not measuring behavior per se, emotions can still provide meaningful insight.
To demonstrate this, Thom Noble (CloudArmy) and Chris Christodoulou (Saddington Baynes) outlined an experiment that was conducted where a virtual car showroom featured the same car but with several backgrounds. Viewer emotions were tracked after looking at the showroom using quick association testing. The outcome was a strength of ‘emotional pull’ for each image and the valence of that emotion. What’s lacking is a measure of real behavior, such as car purchases, as a direct result of seeing one of the virtual showroom backgrounds.
The context in which actions take place plays a major role in behavior. Resultantly, when understanding human behavior, context is key. How we interact with and understand the world around us differs by how and where we encounter it. Taking data at face value can be misleading, so it’s crucial that data is examined within context.
It’s becoming more important to consider the role of context in AI as it attempts to predict human behavior. AI relies on a word’s literal meaning. When brazenly applied to human behavior, this can neglect crucial input. Words can have multiple different meanings, and multiple words can mean the same thing.
A vast amount of the meaning in human language comes from the context. Without context, the ‘why’ of behavior can’t be understood. AI uses pattern recognition analysis in its’ attempts to decipher human behavior. Recognizing patterns, however, reveals only blatant associations and furnishes nothing new. It’s important therefore that we use AI as a tool, combined with contextual intelligence.
To ensure that research captures context, we should employ observation. This can be done by:
For example, Katie Hollier (weseethrough) detailed the use of head-strapped cameras to allow observation of hands-free behaviour in the home. This can provide insight into how people naturally interact with their environment. By observing someone in their own home, and from their very point of vision, projection of our own lives is minimized. Moreover, discrete fixed in-home cameras increasingly become ‘out of sight’ and thereby out of mind – allowing measurement of real behavior.
Moreover, in his talk, Bart Muskala (Accrutat.ai) looked at how large quantities of location data is available for observation via people’s mobile phones. This data can assist in building a picture of the context around research findings. For example, which customers shop only in your store and what types of other stores do they visit? Such data can provide the necessary context to target individuals with appropriate ads or deals.
Ad impact can also be assessed by observing passive online behavior before exposure to an advert and then re-examining behavior post-exposure. This gives insight into the advert’s potential influence on category interest, web searches, and shopping behavior. EEG and eye-tracking during exposure can help identify specific aspects of an advert that influence or the effects of cognitive load on the impact of the adverts.
The ‘easy wins’ for behavioral science so far outside of academia have been small tweaks to existing communications that lead to changes in behavior. For example, modifying NHS text reminders to increase attendance rates for appointments.
These communication tweaks often utilize behavioral science principles, such as social proof and loss aversion, but also aim to simply make copy more understandable. Making things easy is the first principle in the Behavioural Insights Team’s EAST framework and is something copywriters are adept at.
When writing communications, Hannah Moffatt (Schwa) detailed several things to consider: