This newsletter issue is the second of a series based on a podcast we did with Beyond Philosophy’s own Zhecho Dobrev (@ZhechoDobrev), author of The Big Miss: How Organizations Overlook the Value of Emotions. We talked with him before on our podcast a month ago about how emotions drive customer behavior. This week, we dive a little deeper into this concept and explore how emotions affect the insights we get from Customer Science.
Customer Science is a relatively new development in the area of customer strategy. It combines data, behavioral science, and artificial intelligence (AI) to gain insight into customer behavior and predicts what they will do next.
On another podcast, we learned about how AI has some shortcomings, but they are humanity’s fault. For example, some experts assert that AI is opinions written in code.
That said, AI technology provides options for predictive analytics in experiences we never had access to before. We can anticipate customers’ needs tomorrow based on their behavior yesterday. Therefore, imperfect or not, we see that if AI is applied intelligently and with a deliberate approach, it has the potential to problem solve and help organizations achieve their customer strategy goals.
One significant area that could use improvement regarding AI is the area of emotions. We know that emotions drive customer behavior. If people build AI systems and miss this, they can have problems.
Dobrev addresses those areas of emotion and AI. AI needs a lot of data and needs to be the correct data. Therefore, the data must include emotional data. Otherwise, the insights you discover with Customer Science may not reveal the authentic drivers of customer behavior. Researchers and Customer Experience champions must be aware of this fact.
Good Examples of Organizations That Get It
Dobrev points out that some organizations already get this concept and are doing an excellent job gaining insights from it. First, he discovered a company through shared interests, conferences, and research articles, called NICE Enlighten AI CX Program, which he covers in the book. One of their clients was a big US bank. They discovered through AI analysis of hundreds of thousands of call center interactions that agent behavior leads to positive customer interactions.
The behavior that led to these positive outcomes echoes Dobrev’s research in the book. Listening to customers, empathizing with customers, and showing respect led to positive results. The AI tool from NICE Enlighten AI helps agents by coaching them to better customer interaction and giving them a sentiment score, which measures the emotional outcome of that interaction, for each call.
Another example of an organization using AI well was a mortgage company in the UK that used a model that predicted which contact center agent might be the best match for an incoming customer. It used the customer’s characteristics and history to match them to the different agent types in the contact center. Therefore, if a customer is more of a thinker, they would go to agent A; an actor might go to agent B, and so on.
Dobrev also liked a Canadian Start Up using AI to predict when an account holder was about to default on their loan payment. The company works with large companies, like banks, telecoms, and the like, to determine through data analysis and behavioral science when a loan holder is likely to defer a payment, which is the first step in a default. Then, the start-up will intervene, nudging the customers to restructure their payments through a self-service menu. Moreover, the AI uses natural language processing (NLP) to determine the sentiment of the customer reaction. The next nudge they send will adapt based on this sentiment. Dobrev appreciates how that creates a more empathetic and proactive customer experience without driving more traffic (and costs) to the contact center. Moreover, it improves customer interaction.
One of the significant problems in experience is replacing people with systems.
While the systems do not cost as much and tend to be more efficient, they lack the flexibility to respond to change. Unfortunately, systems that cannot respond to a need to change transform an efficient and inexpensive experience into a bad Customer Experience in some instances. The start-up in Canada is an excellent example of creating the best of both worlds: an efficient and inexpensive system (compared to human labor) and the ability to respond to change when needed.
Bad Examples Are There, Too
Organizations jump on the AI bandwagon like any new trend without understanding it well enough. However, they think they do. It’s an example of getting insights from a Wicked Learning environment.
Some of you might remember from our podcast on The Myth of Experience that a Wicked Learning environment is where you think you have all the information but don’t. Many organizations believe they have enough data to decide, but they don’t. Then, when these organizations go to build AI systems, they go wonky.
Why? They don’t work because the AI is based upon what the organization thinks they understand about customers.
Overall, the danger is that organizations build their AI-based solely on the assumption that the customer has only rational experiences and leaves out the emotional and behavioral sides of things. Without them, the nudges Zhecho describes in the good examples will not be possible.
It’s an example of the old standby “garbage in, garbage out.” Guess what? If you don’t put anything into the data about emotions? You aren’t getting anything out about feelings—hence, “The Big Miss.”
We believe that AI is the future. We don’t think that AI will replace humans. However, we believe that human managers who use AI will replace those who don’t. Likewise, organizations that use AI to create better customer experiences will replace those that don’t.
However, it all starts with what data you have. Then, you can determine what information you need in the future to create the experience that you want. If we begin to get some of that together today, we can prepare for the greater good for organizations and customer strategy tomorrow.
If you have a business problem that you would like some help with, contact me on LinkedIn or submit your pickle here. We would be glad to hear from you and help you with your challenges.
