We have all likely had a poor customer experience that in some way, shape, or form involved artificial intelligence (AI). One example familiar to many is an interactive voice response (IVR) system asking a caller a series of questions – often to resolve a customer service issue – followed by an interminable wait on hold, followed by a connection to a call center agent who invariably asks the same questions the IVR already captured responses for.
It is no wonder that according to research from Vonage, 61 percent of consumers say that IVR makes for a poor customer experience, with frustration, stress, and anger cited as commonly triggered emotions. If the purpose of AI in the customer experience realm is to facilitate or simplify a human-driven experience – to mediate the intersection of employee and consumer processes – then the IVR example above accomplishes the opposite. Yes, the IVR software captures important customer data, but the duplicative process ignores the customer at the heart of the process, and introduces friction into the customer journey.
AI at Your Service
Done right, the use of AI in a human-based customer interaction can reduce friction, but to make this happen, the right data need to be available, the right scripts or activities must be built into the process, and the AI should help move the process along to a favorable outcome, either through a next-best action or decision, or by learning (and then reusing) an outcome that improves the customer’s level of satisfaction.
In most typical, real-world processes, other issues will crop up. For another example, an encounter with a hotel front desk associate is a common interaction – either face-to-face, over the phone, or even through a chatbot. In this case, a customer data record barely scratches the surface for the information an associate will need to resolve a customer issue. For a room change or to book a spa appointment or tee time, for example, an associate will need access to inventory, availability and pricing, be able to book and update the systems of record, and ensure everything is updated in the customer record to assist both the current customer and future customers who may make similar requests.
In-store clienteling offers a similar data mining exercise that should similarly be frictionless as part of the process of satisfying the customer and offering the customer a seamless, personalized experience. Is a product available in the customer’s size or color preference? Is the associate able to immediately offer five related products that may be of interest to the customer, that are also in the right size and color and currently stocked? If a product is out of stock, is it available at another store or for direct shipment to the customer’s home?
Depending on the use case, AI can enhance a customer experience in multiple ways that should all feel organic to the customer. For each use case, the baseline requirement for incorporating AI into a process is to ensure a tight connection between whatever part of the process AI facilitates, any other systems the process touches, and the humans involved in the process.
This baseline requirement is also true for other interactions that may indirectly involve a customer, such as AI telling a company or brand something about sentiment, intent, or customer status when a customer sends an email, answers a question in a chatbot, or has posted a message on social media.
Add Value, Behind the Scenes
Beyond use cases for deriving intent, gathering data (such as a clienteling app finding similar product recommendations) or having an interaction with a customer (as in IVR), AI more broadly can help with the decision-making part of the process. If we think, for instance, of a front desk associate at an upscale hotel, AI can either make the decision for the associate or generate recommendations that an associate can choose from. Ultimately, the method should be whichever helps reach the desired outcome with the least amount of friction. A customer likely will have little patience for an associate having to bounce between four or five applications on multiple screens, the classic “armchair integration” situation.
In addition to helping with decision-making, AI can also be used to analyze the decision for future human-to-human interactions. Did this specific decision move the customer along in their journey appropriately? Was the customer’s sentiment, stickiness, or loyalty improved by the human-to-human process? In this context, AI can look at every detail of the interaction and put it back into the customer record, to improve both the individual journey and to learn how to enhance future journeys.
At its best, the use of AI in a human-driven interaction may not even be immediately apparent to the customer on the receiving end. A customer may simply see that their issue was expertly resolved, their needs perfectly met, and end up with a deeper appreciation for a brand that always seems to know what matters to them as a customer.