One of the hardest things for marketers to understand is how people feel. This is quite difficult because we don't think about how we feel as we're feeling it. We have been conditioned to give rational reasons for “why” we do things. And for those providing feedback to ads in focus groups or through surveys, people often say what they wish were true, or what they think the surveyor wants to hear.
We interviewed Graham Page, Global Managing Director of Media Analytics at Affectiva. During the interview, he talks to us a bit about his background in media analytics, the challenges marketers face, and how Emotion AI serves as a tool to understand natural response to ad content.
How has your career path taken you to Affectiva?
I've worked in the marketing insights industries for over 25 years, primarily helping brands and advertisers understand how people feel about their brands and also how to optimize their marketing communications to help people make purchasing decisions.
As part of that process, I've spent a lot of time looking at methods from cognitive science and neuroscience and how those methods can help marketers understand people better and offer a stronger return on the investments that they're making.
I did a lot of work looking at how different ways of measuring people's responses might help: putting people in brain scanners or sticking EEG skull caps on them. Those things are interesting, but none of them are particularly scalable or comfortable. So a decade ago I was looking into the utility of people's facial expressions as a means of understanding their responses. The company that I did that with was Affectiva, and so I've worked with the business for many years and that's led me to ultimately lead the media analytics group here.
What are some of the challenges of those who work in the media analytics space?
One of the hardest things for people in media to understand is how people feel, because we don't think about how we feel as we're feeling it. People are almost trained to give rational reasons for why we do things, but actually why we might decide to do something (i.e., buy a particular brand, or make a life choice) is more complex than we realize. There are more instinctive reactions that frame our more considered thought processes. Understanding those instinctive responses is quite hard and yet, increasingly marketers and businesses are realizing that you've got to understand that initial emotional response to really understand why people do something. That is exactly the challenge that we help marketers and businesses address.
Based on your work, what do you recommend brands or marketers focus on when they're thinking about this?
The more that brands and businesses can understand their customers, the more effectively they can create services that people really want and thus the more successful they will be.
Everyone knows that emotion’s role in buying decisions is important, but a lot of metrics that businesses use to make their choices (such as their overarching marketing strategy) are still quite rational. They're looking at hard sales figures and they're looking at consumer feedback via social media, when they should be looking at consumer reactions in the moment, be that to a brand experience or in response to one of their ad campaigns.
Adding this level of detail and insight helps businesses get better at delivering for their customers' needs. As a result, brands over the last decade have started focusing more on sophisticated metrics based on what people feel to inform better business decisions.
We're also working in the automotive space, providing in-cabin sensing solutions to car manufacturers. I understand you may have some insights surrounding automotive and market research?
So, automotive brands rightly spend a lot of their time optimizing the car itself: the in-cabin experience and the features that entails. There's a lot of research done to make sure that people are going to be delighted with the different features. A lot of that takes time and is complicated. Adding in a layer of in-the moment emotional insight in addition to asking questions can help spot when something is going to be a problem. For example, if a new feature will annoy people, or if people are excited about a particular experience.
We've already done work with some car manufacturers where people have been invited to try out a new feature and we've been able to—with their consent—record them and then use that insight to say, "You seem to be struggling here: can you tell us why?" This allows you to hone in on a possible problem and use more conventional research to understand why that was. You can see that there's a lot of opportunity for Emotion AI to take this to the next level.
To build on that, what do you envision as the future in-cabin experience where content plays a key role in personalization of the transportation experience?
As the interior of vehicles become more of an entertainment space and less of a driving space, these sorts of applications become really important. Of course, if we can understand what is lighting people up about the content that they're watching, then that offers a number of opportunities. As a passenger, if the system understands that you would like a particular form of content, (you tend to enjoy comedies rather than serious dramas), that content can be tailored for you as soon as you sit in the vehicle because it knows who you are, and how you tend to respond. So there's means by which we can make the occupant experience better.
Conversely, in the example of backseat taxi advertising, those systems don't really understand how people are responding. If those systems did understand that a certain type of content is not working then the system can switch that out for something more positive and effective.
Where can our listeners go to learn more, or do you have any other call to action for them?
Definitely go to Affectiva.com. Also try the Affectiva Emotion AI Summit site: I gave a talk there on the role of purposeful advertising, and discussed how important that is and how people respond to ads with purpose messages.
We also have the next Emotion AI Summit, scheduled for October 14th, 2020. Given the huge amount of data that we have—about 48,000 ads in our database to date—gathered from all over the world, responding to different types of content, we'll be sharing a lot more about what we’ve learned from that at the next Emotion AI Summit in the Media Analytics track.
Hear more from Graham’s interview in the Affectiva Asks podcast