Service organizations create value by delivering low-effort, high-impact customer interactions. Since every company is unique in their customer experience, sometimes a custom solution is optimal. In this video, learn about options for innovating to create a unique customer experience.
- If you boil it all down, we have two big jobs in customer service: creating great customer experiences and doing so efficiently. Let's focus on how AI and ML can help us create great customer experiences. Before we go too far with that, we probably should say what we mean by a great customer experience. A recent book called "The Effortless Experience" from the Gardner subsidiary CEB looked carefully at thousands of customer service interactions. They wanted to find out what things were best at creating customer loyalty. Somewhat surprisingly, they found that even wow experiences weren't very good at moving the needle on loyalty. Their provocative conclusion was stop trying to delight your customers. On the other hand, their research showed that it was incredibly easy to erode loyalty. The biggest predictor of loyalty erosion was service experiences that took a high degree of effort from the customer: too much time, too much bouncing around, too many times repeating themselves. So another way to frame our question is how can we use technology to reduce customer effort? In my time talking with service executives and practitioners, I've seen some intriguing examples. One thing machine learning algorithms are good at is making recommendations. Taking what we discussed with search engines to extremes, one technical support organization uses five years of data to train an enormous neural net. For each customer interaction over that time period, they give the algorithm the customer's initial question and the knowledge-based article that a highly-trained engineer eventually used to resolve the issue. Maybe months later. The algorithm then recommends a KB article for new customer questions. Despite the complexity of these problems, their success rate is high and increasing, a truly tremendous outcome that can dramatically reduce customer effort and save the company time as well. Wouldn't you like to know where your customers are in their journey with your products and services? If you knew whether they were just starting with it, or if they're power users, or if they're at the risk of churning, you could tailor the service you deliver and maybe even your marketing offerings. ML can cluster like things together, and we've seen ML algorithms that cluster knowledge-based articles with other ones that are all used in the same stage of the customer journey. So when you see what articles customers are looking at, you can get a sense of what kind of user they are. ML algorithms can predict the future. It's true that history repeats itself, and these patterns can be used to train a model to predict what happens next. Well, one thing customers would rather not repeat please is the experience of having one issue, getting that resolved, and then having another question arise, making them have to contact you again. ML can discover that one question is frequently followed by another, so if we can predict that a customer who has one issue is then likely to have this other one next, we can proactively deliver that answer too, avoiding the next call. This lowers customer effort significantly. These are all examples of how ML can improve the customer experience, but they're only the start. So how best can you find customer effort to eliminate? The first way is start measuring it. The Customer Effort Score or CES is an increasingly popular metric that couldn't be simpler. Rather than asking customers if they're satisfied with their interaction, ask them whether they agree that it was low effort. Then seek out the high effort interactions and figure out how to make them easier. I bet you'll find ML is often an effective tool to do that. Another approach is to map the customer's journey with you. Rather than just looking at things from your perspective, build customer empathy by walking through interactions from their perspective. Do they have redundant interactions with different groups? Are they left hanging with no insight into what you're busy doing? Sometimes fixing the customer experience, pain points you uncover is as simple as changing processes or communicating better, but sometimes AI and ML can reduce the effort and frustration you discover through journey mapping too.