Product development in the era of ‘decay of delight’
Often the first goal in customer satisfaction is to ‘surprise and delight.’ Most product managers are not dreaming of a product or feature becoming obsolete. They’re not usually crossing their fingers that users won’t even notice a release or update. So maybe we’re a bit of a different breed, because we revel in the decay of customer delight.
We’re exploring many different technology solutions to customer problems, including augmenting customer service with Pepper the robot and exploring how artificial intelligence (AI) can help make banking better. In 2017, we also launched our virtual banking assistant in Facebook Messenger.
The virtual assistant is an AI-powered consumer facing experience that allows a customer to type in natural language and receive an answer back from the bot. More than just a technology play or flashy product, with the virtual assistant, we aimed to solve a crucial customer problem. People want consistency in service and answers. And they want twenty-four hour access. We were able to provide just that, in a channel that customers can fully control.
For many users, the first time they interact with a bot, the novelty is delightful, as they engage in natural conversation with a machine and are amazed when it provides answers or completes a task. Think of your friend showing you they can turn the lights on and off with a verbal request or change the thermostat without moving a muscle.
Our desire is to deliver this kind of gratifying experience to customers, but we also recognize that when users interact with the virtual assistant, they want to feel as though they are in a conversation with their bank. This comes along with high expectations of what the virtual assistant can accomplish.
With each interaction, the AI-powered chatbot is processing and understanding, using machine learning techniques to get better over time. The end promise is that each time you use it, you are contributing to it getting smarter.
One of the biggest barriers to improvement is getting users to push the system to the limit. We are actually not hitting the edge enough. When people feel the delight of a novel product, or the frustration when it can’t do exactly what they want, they are less likely to push the extreme boundary.
As we dig deeper into natural language processing and how humans interact with robots, we have developed a deep fondness for the natural decay of delight as people become more accustomed to what bots can do.
We want people to constantly increase their expectations of the capabilities of a bot, as odd as that may sound, because it pushes us to continually improve and build new capabilities.
As market leaders, we have this great problem of being at the forefront of people’s expectations. We can give them everything they want, then they’ll want a bit more. That is what drives us forward. We are constantly discovering, and solving, new challenges - which is the most exciting part of what we do.
As a product team, we’ll always work on what has the most customer impact. Business value is customer value. Sometimes that means we’re shipping brand new features. Sometimes it is simply: ‘here is a fast stable product that you can trust to work’.
Updates are not always announcement worthy, but that is actually the goal; people may not notice, remark, or comment, but we want them to have that increase in expectations and decay in delight, because it means we get to keep making the product better.
With the virtual assistant, we aimed to provide core customer and business value. And by no means are we done. The Facebook Messenger virtual assistant was a direct result of research and listening, with an ideal confluence of our internal capabilities, a great partner in Finn.ai, and the right channel with which to integrate. We are continuing to explore the many vectors and areas where we can make the experience better and applying our learnings to other high value services for ATB customers.