The emergence of emotion analytics—the process of recording and analysing emotional responses to ads—offers new ways to measure and quickly adapt marketing messages. Powered by cheap yet sophisticated sensors, cameras and machine learning, it can result in additional insights to traditional market research methods such as surveys and web analytics.
Most significantly, emotion analytics captures valuable System 1 data (people’s immediate, visceral, and emotion-driven responses and way of thinking) that is often difficult to articulate, yet it’s this thinking that is understood to play up to a 90% role in brand affinity and purchasing decisions.
Via real-time remote panels, we can measure the subtleties of micro-expressions, eye movements, heart rate, and breathing—the signifiers of emotions that contribute directly to brand affinity and purchasing decisions. These are important, considering that advertising campaigns with purely emotional content perform twice as well as those with only rational content, according to the IPA Databank study.
A number of emotion analytics firms, such as Realeyes, Affectiva, and Sensum, are now converting this System 1 data into appropriate, actionable metrics. Ultimately, this is empowering marketers to:
Determine What Creative Will Work
Experience and gut feel from ad creators and marketers has always been supported by creative testing. By applying emotion analytics, marketers can eliminate guesswork to understand precisely what parts of their video content engage, as well as exactly where they are losing viewers or eliciting an undesired emotion. Whether looking at a 15-second pre-roll or an hour-long webinar, marketers have a better opportunity to make sure content is seen—and felt—by the right people: a critical factor for video and interactive adverts that are especially costly to produce.
Power Up Programmatic Targeting
Information about audiences’ online behaviours and preferences already fuels most programmatic targeting. Integrating emotion analytics data could add System 1 reactions, or even likely emotional states, to that profile data. In practice, this enables us to deliver the right message at the precise micro moment of emotional warmth and receptivity, avoiding the risk of annoying or inappropriate messaging. On this point, it’s worth noting the nuances of positive and negative emotional responses. Effective stories deliver an emotional arc, whereby sadness, confusion, or concern in subjects doesn’t mean an ad isn’t working—in fact, it can mean quite the opposite.
As a rule of thumb, a highly engaged audience will reflect the themes of the creative: expressions of concern or disgust to an ad portraying germs that need to be killed by a cleaning product would be considered desired responses.
Deliver Ultra-Precise Creative
Programmatic advertising requires ever more variations of adaptive and quality creative. Many marketers struggle to produce and manage the volume of creative variants required. While behavioural analytics and dynamic content optimisation (DCO) now increasingly decide what creative appears on screens, emotion analytics data can help inform which delivers the best impression, thus eliminating waste, driving efficiency, and powering more meaningful engagements. Provider Realeyes builds a compelling case for the use of emotion tests that can predict long-term sales impact with up to 75% accuracy, particularly important for high-cost, high-stakes environments where there’s no room for error.
At this nascent stage, emotion analytics requires ethical and moral considerations. Our online identity and behaviour are increasingly being scrutinised for marketing purposes, and issues of privacy are already high on the agenda. The upcoming GDPR (General Data Protection Regulation) rules will require companies to take this issue far more seriously in the coming months.
For marketers, the constant core principles relating to any data-led technology remain front of mind. We have to be transparent, making clear upfront what we’re measuring, why, and how users’ data will be applied. There has to be a clear value exchange, whereby users have an incentive to opt in and allow greater measurement of their information. This could mean offering improvements to user experience, or an attractive novelty factor.
Finally, marketers will need to be flexible enough to allow for alternative types of value exchange, perhaps with a survey, or a phone call using vocal analysis. As with all preceding data advances, consumers come first.
Marketers have long had access to technology that helps them measure and predict emotional responses to content. Recent developments in sensor technologies, however, and the proliferation of smart media devices and cloud computing services have moved this on from the unwieldy, expensive realm of R&D and experimentation to practical, accurate indicator of actual marketing success. For those wanting to understand how to target their efforts more confidently and precisely, while delivering great results in the long term, emotion analytics is an increasingly viable option.