Empathy, expertise and ethics: Delighting customers via augmented AI

When was the last time you walked away from a web, app or phone experience with a service provider, raving about how much you enjoyed the experience?  If you have, you’re one of the lucky few. Unfortunately, awesome customer experiences are hard to come by. Let’s see what the data says. In an oft-cited study

When was the last time you walked away from a web, app or phone experience with a service provider, raving about how much you enjoyed the experience?  If you have, you’re one of the lucky few. Unfortunately, awesome customer experiences are hard to come by.

Let’s see what the data says.

In an oft-cited study a few years ago, Bain & Company reported that 80 percent of B2C companies believe they deliver a “superior experience” to customers. Yet, only 8 percent of customers agree! And, it’s not much better for B2B buyers. More than 80 percent of business purchasers surveyed by Salesforce expect a consumer-level customer experience – and two-thirds have switched vendors to get it.

How can that be? It’s because transactional, one-size-fits-all approaches no longer work in today’s marketplace. Customers are looking for deep, personalized relationships that span not just one but multiple transactions over time. Over time, such interactions can build deep trust between the customer and service provider.

With the advent of AI-driven smart assistants, what was once a rare, costly and near-impossible proposition now has a chance of becoming a reality in the coming years.

Let’s take a look at what the future may hold for customers connecting to financial services, a domain we’re focused on every day here at Intuit. This video imagines an awesome customer service experience between John and Elly, an AI-powered voice assistant with deep expertise in finance.

Here we see Elly providing not only transactional data—the dollars and cents of various options—but also a relational experience. She expresses understanding of John’s situation in all its delicacy. While the substance of her advice could easily have been delivered in a just-the-facts manner, even as a simple spreadsheet, her more nuanced approach makes a huge difference for John in this difficult time.

As he ends the session, John has the information he needs to take the right action, but it’s the feeling of the interaction that will stay with him the longest. Elly was there for him when he needed help, and made him feel confident that things would work out.

Closing the gap to deliver on customer expectations

The first question to consider is how—or whether—AI and human intelligence should coexist. Rather than seeing AI as a replacement for human intelligence, we at Intuit envision a collaborative world driven by AI that smartly augments the human. By melding the fast, accurate information processing of computers with human adaptability to changing circumstances, we can leverage their respective strengths. Ultimately, this will elevate the role of the human expert, eliminating much of the drudgery associated with tedious, manual work.

Building such an experience presents considerable challenges, with hundreds of problems to be solved to deliver a truly delightful customer experience that can be broadly categorized into three core capabilities: empathy, expertise, and ethics.

Empathy: You understand me

Empathy is the most challenging and complex factor in customer interaction. The agent, whether human or bot, must listen carefully, assess accurately, and take control gracefully. The conversation must be appropriately paced, and sped up or slowed down, in keeping with the content. All of which is a matter of judgment, a characteristic that comes naturally to most humans but must be embedded into intelligent machines to achieve a satisfying result.

Giving machines a human-like capacity for empathy will depend on continued advancement of natural language processing and generation, and the ability to read cues such as gestures, touch, and tone. Ultimately, we envision a time when intelligent bots are likeable, attentive, and adaptable—like humans at their best.

Expertise: You can solve my problem

Having a pleasant, natural-sounding conversation with an intelligent chatbot is of little value if it doesn’t solve your problem. Expertise matters. Progress in this area is accelerating, with huge investments by universities, research labs, start-ups and large companies alike. At Intuit, we’ve made significant inroads, using domain-specific finance and compliance rules to build knowledge graphs, and investing in real-time machine learning algorithms for cash flow prediction, accounting and tax prep automation and customer care, for example. And, of course, clean data pipes with scalable information streaming.

So, how do we bring all this expert system capability to bear on the customer experience? It begins with domain knowledge. We build customer confidence by demonstrating a keen understanding of the problem itself and its context, using historical data (with our customers’ permission) to anticipate questions, deliver personalized answers, and make informed suggestions. Above all, we know when to shift gears and to enlist a human domain expert to handle questions AI alone cannot solve.

Ethics: You are trustworthy

Ethical behavior can be as challenging for machines as it is for their human masters. Fortunately, the ethical dimension of customer experience can be narrowed to a few key principles.  Be a trusted data steward, instill confidence in the accuracy of your recommendations and predictions, and practice transparency by giving users a peek “under the hood.”

For example, we built the ExplainWhy feature into our QuickBooks Self Employed product, a self-service interactive dialog that allows users to drill down as far as needed to be confident about our quarterly estimated tax calculations—in just seconds. In essence, we learned the importance of “showing our work” for some customers.  Unlike simple look-up systems, ExplainWhy interacts with the QuickBooks Self-Employed user via an incremental and interactive dialog.

As an industry, we are just beginning to appreciate the importance of embedding human values of empathy, expertise, and ethics into our intelligent technology. The task is challenging, but the payoff will be well worth the effort. Before long, Intuit foresees a time when the humanness of intelligent agents becomes the next big competitive differentiator—and a driver of real, enduring delight for our customers.