One of the best ways to sell products is to engage the customer on an emotional level. Mad Men's Don Draper knew that. (Remember the Kodak Carousel pitch?) If you can tell an emotionally engaging story about the products you are trying to sell, the sales will follow. But, how do you know whether or not shoppers like what they hear?

Emotion AI technology may be able to answer that question, with its potential to apply computer vision, facial recognition and emotion analytics to help retailers and brands understand how shoppers really feel about what they are buying — or not buying.

How it works

Emotion AI systems use small cameras or sensors, deployed in stores in high-traffic locations, such as shelves, aisle end caps, checkout lines, and entrances and exits — anywhere and at any stage in the shopping process where retailers and brands are most interested in gauging shopper sentiment. The system records detect the facial expressions of individual shoppers, and the computer vision, AI and analytics component recognize, classify and interpret the emotions they appear to be expressing at a given moment.

The result could be a compilation of various emotional responses retailers and brands could access to help them make decisions about important factors that may fuel the buying decision, such as:

• Product pricing, packaging and branding: If emotional responses skew negative, it might be time to lower prices. If shoppers examine packages and appear confused, it might be time for a redesign.
• Inventory availability and replenishment: If shoppers appear happy about certain flavors, perhaps stock more of those, less of the flavors they frown at and push aside.
• Shelf placement, store and aisle layouts: If emotional responses suggest frustration, perhaps a reconfiguration of product displays or a move to a different aisle is in order.

An array of technology companies — Affectiva, Conduent, Anodot, NTechLab, Microsoft, Apple, Google and others — are working on Emotion AI technology schemes with such goals in mind.

Given the potential benefits to retailers and brands, market hopes are high. A recent Orbis Research study suggested emotion detection and recognition technologies alone could represent a $24 billion market globally by 2020, and HTF Market Intelligence has said the worldwide market for emotion analytics technologies will experience a compound annual growth rate of almost 83% between 2016 and 2022.

Consumer hesitancy?

With the ability to measure something once thought to be intangible, and the difference it could make in giving customers what they want in terms of product presentation, price, selection and store design, Emotion AI could hold tremendous value for retailers and shoppers alike. The final hurdle to clear is customer acceptance.

Emotion AI technology developers are aware of the potential privacy concerns, and are quick to point out that these systems do not identify individual shoppers, or record their faces and other personal details, instead only recognizing and classifying actual expressions, keeping the data samples anonymous.

Still, this is an evolution that can't be rushed. If Emotion AI enters live shopping environments before companies adequately determine that consumers are ready, the betray of trust could damage the emotional bond they are trying to develop. For example, shoppers could feel they are being stalked and simply stay away from stores that use this technology, or they could even show their resentment by trying to game the system with false reactions (or maybe just googly eyes).

As retailers and brands come to more fully understand Emotion AI, and take time to cautiously test and evaluate it, the technology could play an important role in giving customers exactly what they want. It could help companies tell stories that emotionally engage consumers in the best possible way because their customers' own emotional responses have helped to write those stories. But first, the sector needs to be sure shoppers are ready to collaborate