From customer-facing chatbots to email tagging, AI tools are becoming an increasingly important factor in providing a first-class customer experience. The application of AI technologies is not only enhancing customer interactions but is also yielding up new -- and often unexpected -- information about customers that allows companies to fine-tune their services and products.
Here are some ways in which businesses use AI to improve customer experience.
1. The human-machine dialogue
By the end of 2020, 85% of customer interactions will be human to machine, with chatbots making up a large portion of those, according to Gartner research.
Chatbots bolster the overall support capability of the business. Digital service agents allow business to make the support desk available 24/7. The increased availability increases customer loyalty to the brand. And chatbots respond instantly, so the customer isn't put on hold.
Here are some ways in which AI-powered chatbots interact with customers:
- Data collectors. It's easier for a customer to answer a chatbot's information-gathering question during an exchange that the customer cares about than it is to get them to take a quick survey after the fact. This data bolsters customer-centric analytics.
- Access to business systems. AI-enabled chatbots can tap into enterprise systems -- the customer data platform, in particular -- for immediate access to predictive insights that might serve the interaction well. Once a chatbot determines a consumer's specific need, it can sift through the data of that person's past interactions and preferences and offer AI-based decision support, personalizing the customer experience.
- Multilingual support. One of the biggest struggles that customer service desks face is maintaining a multilingual presence -- it's expensive and often inadequate to meet demand. With AI-powered chatbots, this problem goes away. The chatbot can detect different languages and immediately respond in that tongue, which further develops the chatbot's vocabulary and widens the audience to which it can provide customer service.
- Medical interactions. Chatbots can make medical interactions more effective. Ellie, a chatbot created by a team at the University of Southern California's Institute of Creative Technologies, can respond to not only a patient's words, but also to their facial expressions and gestures, via video camera, and respond accordingly. Therapy patients of all kinds often respond more candidly and completely with a chatbot than an actual human being, because there is no danger that the chatbot might shame them or judge them, according to the Journal of the American Medical Association and a RAND Center study.
2. Human-to-human, AI-augmented
Chatbots are the tip of the customer interaction iceberg, where AI is concerned. But not all enterprise-customer contacts are chatbot-appropriate. Chatbots can't do everything. They basically fall under the 80/20 rule. In customer support, for instance, 80% of the total volume of calls consist of common, chatbot-ready questions -- such as resetting passwords and saving reservation information -- while the remaining 20% require human attention.
When the customer interaction requires a human touch, however, AI can still lend a helpful hand, improving interactions in ways that might not be obvious. One example is email tagging -- scanning customer requests and issues arriving via email to ensure they are routed to the correct human agent the first time, rather than being passed from person to person while the customer waits, leaving the customer unattended.
Another example is when a chatbot determines that a customer's problem doesn't fall within its purview. Using AI-augmented messaging, the chatbot finds that a human agent needs to handle the issue. A customer can use the keywords "frustrated," "stupid" or "help," which indicates to the chatbot that it's time to transfer the call to a human agent.
3. AI on the retail sales floor
Sometimes the effect of AI on the customer experience is completely passive. The customer may not be consciously aware of the interaction, but it's there.
The guided experience in retail is one such interaction. Through the analysis of customer behavior, based on patterns of movement in the store, AI can optimize the placement of products and floor layout. In real time, strategically placed beacons note a customer's approach and send coupons or specials to the customer's phone. In the clothing department, a smart mirror not only displays a customer's image wearing a garment they don't actually have on but upsells matching accessories -- doing the suggestive selling that used to require a human assistant. Not only do customers get through more potential outfits faster, but AI optimizes their alternatives -- and the pressure to buy that might emerge in a human-to-human interaction vanishes.