Writing on his recent research in Forbes on how banks use chatbots, Cornerstone Advisor’s Director of Research, Ron Shelvin shares his namesake rule: “Shelvin’s law states: For every statistic that proves a point, there are two that refute it. Nowhere is this more true than with chatbots.”
He points to two competing studies in particular. One that concludes 86% of consumers prefer to interact with a human agent. And another that affirms 74% prefer to interact with a chatbot while looking for answers to simple questions. These competing stats all add up to this: Only 13% of the mid-sized banks and credit unions surveyed said they have deployed chatbots.
Three mistakes retail banks and credit unions should avoid when deploying chatbots
Too often we find that chatbot projects fall short of expectations for three reasons — all of which foster doubt about their value. This skepticism could be holding deployments back, even though, digital transformation is high on banks’ agendas and chatbots are foundational technologies that banks must implement to power their digital business models.
What mistakes do organizations make in chatbot deployments? We find the three most common are:
- The bot tries to do too much
- The bot is launched and left to its own devices
- No one is charged with ensuring the chatbot experience evolves with customer needs
Chatbots bring value to the business by improving a process where the customer is experiencing friction. By starting chatbot projects that address a targeted business problem, banks can fine-tune the bot to ensure it’s providing the assistance needed while building trust with the customer. As the chatbot gains traction it, in turn, builds value for the business and provides a stronger case for expanding the technology.
Talking about Eastern Bank’s chatbot success, SVP, Director of Customer Service Center, Heather Allen proclaimed that “People are thinking about bots being all-encompassing, life-changing, corporate changing, ‘this is going to change the face of everything’. But, if you think about what’s really taking up time, creating friction for employees, creating friction for the customer, those customer journeys are what you go after. Stop trying to be the most innovative person in the universe and start with what really helps your customers.”
How to build a chatbot that isn’t just smart, but actually has common sense
Start small, be specific. Zero in on a function, a process within that function, and a part of that process that could benefit from a chatbot. Focus on meaningful, high-value use cases that improve areas where customers or employees are experiencing the most friction with a process. This is often around customer service. For instance, in a live poll conducted during a retail banking webinar with media giant Informa, participants ranked customer service inquiries as a function where a chatbot could bring the most value. And reducing contact center traffic by resolving basic inquiries was ranked as the top process within that function where chatbots could help.
Eastern Bank is an example of exactly how to do this right. The largest mutual retail bank in the United States deployed chatbots to help customers navigate and find answers themselves on the bank’s website. Some 17% of customers’ journeys were handled by the chatbot — without the frustration of waiting on hold to get a quick question resolved.
Stay with the bot. A well-designed, intelligent chatbot is essential from the outset. But, one of the reasons that companies fail to realize value with a chatbot is that they spin it up and move on — leaving no one to mind it, to help it grow and evolve. The wonderful thing about artificial intelligence is its ability to evolve quickly as new information becomes available. But it can’t do it so effectively without an established feedback loop and a dedicated person, or team, to drive the next iteration or expansion of use.
Success with small, discrete use cases builds confidence and demand for the use of chatbots. They can move from supporting roles, to being the engine for transactions, to further engaging customers, to being the glue holding different channels together as research from Informa shows. “This is an iterative process. If someone doesn’t own it, monitor it, and keep in touch about trends, it can get stale,” Eastern Bank’s Allen said.
Show customers, they can trust you. One of the best ways to build trust in using a chatbot is for the chatbot to know when it’s out of its league. A well-designed chatbot can seamlessly hand off a customer to a live agent while retaining all of the information about the customer’s problem up to that point. This builds confidence with the customer that the technology is actually working to solve their problems. Trust allows the bots to do more jobs, and deliver more efficiencies and insight for your organization.
See how you stack up against retailers, not other banks. When asked who they look to for examples of banks innovatively using technology, Eastern Bank’s Allen said she isn’t. “I’m not really being compared to competitors. I think we’re being compared to digital retail players,” she said, preferencing that answer with the fact that banks are so different at scale. Consider that customers’ impressions of your digital experience aren’t necessarily being weighed against their experience with other banks. They’re thinking of the interactions they have with retailers with rich digital capabilities. The added benefit of this approach is that it shifts focus to how the banks can use the data that the chatbot is gleaning in a secure and compliant manner. For instance, to personalize offers, to identify and target opportunities for upselling, and to better serve customers.
Building a commonsense chatbot requires thinking small, as Allen says, “Don’t think of what the coolest thing you can launch is. Just think about what the most useful thing is. You’ll be more disruptive by being tone-deaf, by launching this super cool thing when the customer is saying, ‘I just want to replace a lost debit card.”
Are you ready to leverage intelligent chatbots as an essential step in building a digital customer experience? Read our essential eBook — Mastering Digital Customer Engagement for Customer-Centric Teams.