AI-powered chatbots don’t exactly have the best reputation for solving problems for customers in any industry. But this perception may be changing according to a new report from consumer intelligence firm, J.D. Power.
Through its new ‘Lending Intelligence‘ report, J.D. Power further explores its 2024 ‘US Mortgage Servicer Satisfaction Study‘, highlighting the prevailing sentiment and emerging trends in AI-powered customer service, and how that may change with the continued uptick in servicer adoption.
In most cases, early iterations of chatbots left many customers feeling like they were simply wasting their time. But as the technology powering them becomes more sophisticated, their importance could quickly grow.
In the past 12 months, J.D. Power found that 21 per cent of mortgage servicing customers had experienced a problem. Just nine per cent of these used an online chat as their first point of contact, while 48 per cent called customer service instead.
But this may not always be the case. In fact, customers from Generation Y and Z appear three times more likely to use an online chat than older generations – pointing to the fact that this channel is set to become increasingly important.
Already, 83 per cent of customers using the chat to solve a problem said it was resolved using the chat alone. Unsurprisingly, those who were able to solve their problem via chat had an overall customer satisfaction rating of 702 (on a 1,000-point scale) versus 482 for those who could not solve their problem.
Preferring a genuine ‘human’ experience
Seventy-three per cent of customers who used chat say they interacted with a live representative, while just 10 per cent thought it was a chatbot, and 17 per cent were unsure. Those who said they interacted with a human had a better experience than those who thought it was a machine on the other side.
However, 63 per cent of chat users working with a human felt the chat rep used a script, while 37 per cent did not. Customer satisfaction for those who felt no script was used was 699, considerably higher than the average 636 satisfaction score among customers who thought the representative used a script.
Nearly three-fourths (73 per cent) of customers who felt no script was used said that the process was extremely easy – significantly higher than the 27 per cent of those who felt a script was used.
These findings highlight the fact that any investment in AI needs to follow a clear understanding of what the customer wants in terms of service and problem resolution. Without this, J.D. Power says that customers may simply refuse to engage with AI chatbots, leaving mortgage lenders on the hook for the time and resources spent on underutilised technology. Those who can thread this needle will see higher customer satisfaction scores, improved processes, and streamlined costs.
