• United States



by Martin Illsley

Caring about Customers: Emotionally Intelligent Applications

Jan 15, 20047 mins
CSO and CISOData and Information Security

The ability to know customers intimately, to develop strong and deep ties with them, and to use that knowledge to develop better ways of serving them, is a hallmark of the high-performance business. Indeed, Accenture’s ongoing research into the characteristics of high-performing businesses indicates they continue to win the ongoing battle for customers by achieving a deep understanding of their customer’s behaviors and preferences.

Today, increasingly sophisticated technologies and tools are coming onto the market that can improve a company’s ability to form insights and thereby enhance the quality of the customer relationship. But what about technologies and applications that have the ability to understand a customer as more than a collection of transactions, purchases and order numbers? What about an application that could sense, by itself, the emotional state of a customer?

Farfetched? Science fiction? Not anymore. Accenture Technology Labs researchers are investigating the next frontier of computing in which applications are smart in the sense of being emotionally intelligent.

Emotionally intelligent computers

In recent years, the concept of emotional intelligence has escaped from university psychology departments and into the business and popular press. Although the definition of the term may vary slightly from source to source, it refers basically to the ability to influence and shape human behavior, through the awareness and management of emotions. In human beings, emotional intelligence breaks down into a range of interpersonal and intrapersonal capabilities, including sensitivity to others’ emotional states, understanding how one’s personal presentation influences others, and the ability to generate certain emotions and behaviors in others.

Can computers demonstrate similar capabilities? Can they learn to respond not just to a pre-defined set of inputs (keyboard strokes, mouse movements or speech) but to the more complex range of signals people send out all the time-signals that are traceable ultimately to their emotional state?

While the business applications of emotionally aware computing are at least five years from maturity, Accenture Technology Labs believes that these capabilities are a logical stage in the development of context-sensitive services businesses provide to consumers based on their unique characteristics: their physical location, preferences, buying habits and, yes, even their current emotional state.

How is this possible?

How can a machine demonstrate emotional intelligence? Begin by thinking about how you sense the emotional state of another human being. The voice provides one set of cues. Even on the phone we’re able to tell quite a bit about how someone is feeling. In person, we sense a variety of other things: someone may smile or frown, may squint or raise the eyebrows. Gestures tell us a lot, too. We shrug, put our hands to our eyes, scratch our chins.

With the right kind of associated technologies, computers and applications now have the potential to sense many of these same things and make inferences not too different from the guesses we make about the emotional states of those around us.

Voice recognition technology, for example, is not just a new kind of interface or input device. It also can be adapted to make reasonable guesses about people’s emotional states. Face recognition software and software that recognizes gestures through visual data supplied by cameras also are becoming more sophisticated. If you are blinking and yawning, for example, an application could infer that you are tired and might suggest you go have a strong cup of coffee. Mouth pursed and eyes narrowed? You’re probably mad about something. The application might, with your approval, stream soothing music into your headphones. Similar software also can learn and then process information about gestures. As technologies improve, we also are seeing interfaces sensitive to tactile cues, and even to smell and taste.

Emotionally intelligent computers in action

While interface technologies can make a computer emotionally aware, the next necessary step is to program a system so it can respond intelligently based on that input, and help to shape some ongoing interaction with the person. With that capability, how might these emotionally intelligent applications or systems make a difference to people’s lives?

  • When customers interact primarily with a machine. Increasingly, technology is becoming the face of business for customers; how can businesses make that a friendly and caring face? Consider how emotionally intelligent company representatives are able to respond to customers with such phrases as, “You seem to have had some frustrating experiences; what can I do to help?” Computers, on the other hand, tend to say things like, “You are not permitted to perform that action,” or “You have exceeded your daily withdrawal limit.” Computer applications that can be aware of emotional states, and respond intelligently, could influence a customer’s loyalty to a company.
  • Emotionally intelligent computers could encompass elements of human caring. For instance, an emotionally dumb computer cannot sense whether or not an older person has understood information or even, given age-related hearing loss, heard it in the first place. Consider how we talk to older people in a caring manner, sensing their mood and whether they understand and, in fact, have heard the information. Now consider a typical computer spitting out “Take your medicine” or the heart stopping “Intruder alert!” messages.
  • Finally, emotionally intelligent systems could, over time, enable companies to provide innovative and profitable services based on new technologies. For example, an emotionally aware system might work to help the world’s aging population live independently for longer. One project in the planning stages at Accenture Technology Labs creates what we call a “caring plant” in the home of a senior-let’s call her “Mrs. Smith.” Cameras, sensors, voice recognition and production software and other capabilities would be embedded in an actual houseplant that Mrs. Smith tends to daily.

The technologies are unobtrusive and easy to use. The plant monitors its own state, reminding Mrs. Smith when to water it, give it fertilizer and so forth. It also communicates: engaging her in conversation about news, weather and other events in the world and in her life. Through the use of video monitoring and speech analysis, the plant makes inferences about the Mrs. Smith’s emotional state, and can interact with her with, for instance, gentle reminders to take her medicine, or alert a family member or medical professional if needed. By using emotionally intelligent systems, the caring plant becomes a constant source of help, aiding Mrs. Smith through some daily difficulties.

Vision versus reality

Although emotionally intelligent applications are visionary, all the technologies needed to make them a reality already exist. The primary barrier to the take-up of emotionally aware applications and computers capabilities, at least for now, is the fear of intrusiveness. But consider that millions of people around the world already pay an outside security service to monitor their home through sensors and alarms and it is not too difficult to see consumers welcoming the type of “well being” in-home services described above.

Furthermore, these kinds of applications and services that sense and learn things beyond what we tell them need not make us afraid (we can always shut them off, after all). Applications today try to anticipate things people need by correlating all sorts of data: location, current activity, transaction history. Adding emotions to the mix is not as far-fetched as it may at first appear to be.

Before the current decade is over, Accenture believes every industry will sense the potential opportunities and attendant responsibilities associated with emotionally aware computing. There is no time like the present to begin the discussion.

Martin Illsley, director of research-Accenture Technology Labs, is based in Sophia Antipolis, France. He can be reached at

Accenture Technology Labs, the technology research and development organization within Accenture, has a 16-year track record of turning technology innovation into business results. The Labs create a vision of how technology will shape the future and invent the next wave of cutting-edge business solutions. Working closely with Accenture’s global network of specialists, Accenture Technology Labs helps clients innovate for competitive advantage. Labs are located in Chicago, Illinois; Palo Alto, California; and Sophia Antipolis, France.

This article originally appeared as an Outlook Point of View, an Accenture publication. © 2003 Accenture. Reprinted by permission.