Using AI knowledge for a no-effort experience
At my local takeaway, I order the same thing every week. The staff never ask what I want. They simply say, “Mixed vegetable fried rice?”, take the container I pass over the counter and return it filled with steaming goodness. Years ago, when I lived in London, I regularly visited a bookshop tucked away in a court off the Charing Cross Road. Through repeated purchases, the manager grew familiar with my obscure collecting habits. When, as part of a job lot, he acquired a vintage magazine he thought I would like, he saved it behind his counter and threw it in with my next purchase for free.
The customer experience involving the lowest possible effort is one in which you predict what the customer wants, before they even ask.
This kind of customer experience is uncommon, which is why it’s memorable. It requires a level of familiarity we don’t achieve with many of the organizations with which we interact. It requires knowledge of our habits, purchasing patterns and history. For that reason, we might associate it with smaller, local businesses with which we have a personal relationship, rather than with big enterprises.
Yet large enterprises can replicate this experience given the right data and technology. By exploiting the wealth of customer data held in computer systems, and interpreting that data through AI, you can give a self-service bot or busy contact centre agent the same knowledge of the customer my friend at the bookshop had.
AI That Remembers
Let’s take a typical customer service scenario. Last week, Mira bought a new washing machine. The machine isn’t working properly, and she’s not sure if it has a fault, or if she’s just using the wrong settings. She visits the manufacturer’s online self-service. The self-service knowledgebase holds lots of information about troubleshooting washing machines. There’s so much information, in fact, and for so many different models of machine, that if Mira had to find her answer through search she might get confused and discouraged.
Fortunately, the site has intelligent self-service that actively delivers content before the customer inputs an action. The AI, with access to sales data, knows Mira recently bought a washing machine. It also knows which model she bought. It thinks to itself – figuratively – “Ah-ha! She’s probably here about the new washing machine.” And it automatically suggests knowledge articles about troubleshooting Mira’s model of machine, enabling her to find what she needs with minimal effort – while keeping open the option to search the knowledgebase in case its intuition is wrong.
This online experience is effectively the same as Mira walking into a store, and meeting a salesperson who remembers her and what she bought.
Personalizing the Customer Experience
An AI, with access to records of sales and service transactions over long periods of time and multiple areas of a business, can potentially know more about a customer than even the most observant human representative. This enables every service experience to be personalized. The AI delivers content tailored to what the customer has previously bought or done, anticipating the most likely needs from its knowledge of past behaviour. Drawing on existing data, it can suggest help as soon as the customer lands on a self-service page, before the customer makes an explicit request by running a search or filling in a form.
This approach requires:
- Centralized information about the customer’s previous transactions, e.g. from sales records, CRM, and server logs.
- An AI that can make logical deductions from the available data and serve content the customer is likely to find helpful.
- Leaving alternative routes to information open, in case the AI suggests something unhelpful or lacks sufficient data about the customer.
Many sites already use personalization techniques to push products the customer might want to buy. It needs only a shift in thinking to apply the same principle to self-service.
At Transversal, we call this “zero-click knowledge”. The customer receives the information they need without needing to perform an action. This is in tune with the current interest in effort as a determinant of customer experience. It’s not a solution to all transactions, and for some businesses it’s still more of an ideal than a reality. But it offers a powerful tool for making anonymous customer service feel as responsive and familiar as a personal relationship.