Duane Varan, CEO of MediaScience – a neuroscience research firm that worked on this study with the Ehrenberg-Bass Institute – discussed this topic at the Advertising Research Foundation’s (ARF) 2017 Audience Measurement conference.
More specifically, their analysis incorporated over 100 advertisements for 20 brands owned by Mars Inc., as well as laboratory testing from a panel topping 1,000 respondents.
In determining which metrics yielded the greatest insight into consumer attention, the investigation explored indices like facial coding measures, biometric measures, skin conductivity, heart rate, blink rate and visual fixation.
“We were really looking for something that could scale,” Varan said. (For more details, read WARC’s exclusive report: How to measure the neuroscientific power of an advertising heartbeat.)
And heart rate, he asserted, was a very useful readout. “What we are looking for is cardiac deceleration,” Varan said. “When you pay attention to new stimuli, you are basically pulling resources from your body to attend to that.”
Cardiac acceleration as an expression of increased emotion and, presumably, interest might seem counter-intuitive, Varan conceded. But the interbeat interval proved valuable when tracking several aspects of creative department.
Voiceovers are only one example. This feature, in fact, was commonly associated with weak-performing ads in the dataset, as 50.7% of “weak” ads included voiceovers, compared to 29.1% of other ads.
“When we look at this data,” Varan told the ARF audience, “we can actually discriminate using heart-rate data between a good voiceover and a bad voiceover. That's the kind of utility that we end up with.”
Such learnings provide the types of insights that don’t just indicate whether voiceovers grab attention but, more precisely, hint at which voices work in what kind of setting.
Findings in areas such as messaging complexity and the presence of celebrities in advertising further validated the research technique used by MediaScience and the Ehrenberg-Bass Institute.
“This is the ultimate kind of practical application for this body of research,” Varan offered. “This is where you want to end up: With process measures you can measure second by second.”
Data sourced from WARC