Bots and Humans: Use the Best of Both in Your Customer Experience Strategy

We have a joke in my family about eating out–if the service is slow, get me talking. As soon as I begin to speak, our server will arrive with the specials of the day, or a dessert offer, or (my favorite) an inquiry about how our experience is going. Well, fine except for you interrupting me, is what I want to say!

As I think about how AI is currently working to improve customer experience, I ask myself: In this scenario, would a bot be better than a human?

In this case, the answer is ‘no’. At a nice restaurant, I prefer a person who can better read my expression, perceive the mood at the table, and see if the food is rapidly disappearing, or maybe not going over so well. But, clearly, timing and efficiency also matter. Imagine my reaction, for instance, if our waiter brought my side of ranch to the table after I’d finished my food. There’s no doubt that the right (or wrong) interaction at just the right (or wrong) moment is part of the overall customer service experience, and this will play a huge part in how I recall the restaurant. Moreover, it will be a big influencer in whether I recommend it or go back.

Which brings us to how bots can fit into a successful customer service strategy.

In retail, bots hit the sweet spot when it comes to helping us consumers solve our own problems—pre- and post-purchase, for example, research shows that we don’t necessary want to overcomplicate matters by bringing in real people. In a recent Narvar consumer survey, 98% of online shoppers said they prefer speaking to a person rather than a bot, but 38% couldn’t tell if a live chat or messenger app interaction was with AI or a human. On top of that, only 10% could confidently tell that they weren’t communicating with a human.

When answering simple questions or helping consumers with basic tasks, bots are a great first touchpoint, and may even be better than a human representative when all a customer wants is to get an answer and get on with the day. With my side of ranch, for instance, it would have been a better experience for me to simply put in a request at the exact time I needed it, and not wait around for the server to bring it on his own time.

The important thing for retailers is knowing when and how to use your people, and when to save resources by using a bot. The best strategy may be a judicious blend of bots and people, with a sensitivity (that human quality!) toward the right timing and channel through which to deliver the right communication. Based on the results found in Narvar’s consumer study, Bots, Text and Voice: What Cuts Through the Clutter, here are some suggestions for finding that balance:

Use bots where they add more value than a human interaction can.

Bots are great for assisting consumers when they seek answers to basic questions, such as about an EDD (estimated delivery date) or the specifics of your return policy. They can also be used to escalate more complex problems to the appropriate human, driving more efficiency across the board for customer service teams. Bots can also be a great tool for gathering feedback by asking customers about their overall satisfaction with their experiences.

Use humans where they count most.

Much as you’d expect, humans are much better than bots at solving complex problems and handling emotionally charged or difficult situations (the words “I hear you,” for example, can go a long way). It’s also a good idea to involve a human touch when you’re dealing with higher-end products and experiences—84% of consumers say that more communication is critical if a purchase is expensive, and while that doesn’t exclusively mean the communication needs to come from a person, you want to have someone on-hand and prepared in case an issue needs to be escalated.

Create the bot-human workflow with intention—and mind the gaps.

Make it easy for a bot to escalate an issue—you don’t want to frustrate your customers further with interactions that create more friction within their overall experience—and make sure the information your customer has already given to the bot gets to the service rep once the issue is escalated. Keep in mind, too: while emotionally intelligent bots aren’t a reality quite yet, they’re currently being designed to recognize signs of frustration, which will make it easier for them to know when to bring a human into the equation.

The important thing for retailers is knowing when and how to use your people, and when to save resources by using a bot.

Conversational bots and other AI-driven channels such as voice may be in their infancy today, but as they begin to evolve and feel like human interactions, they’ll surely become the norm. Bots are great at assisting with easy Q&As; we like to take care of ourselves until we can’t. When problems need to escalate, then, routing to a human is the right choice.

As with any new communication channel, it’s important to experiment, analyze, refine, and iterate. Just as my family and I appreciate a waiter’s good timing when eating out, shoppers will appreciate bots if you use them to add more value to your pre- and post-purchase experiences.