The banking and finance sector is ablaze with chatbots. Bank of America, HSBC, Wells Fargo, Swedbank, Capital One – these are just a few examples of major banks that have (or are set to) release customer service chatbots.
For banks, bots are a cost saving; for customers, they are a simple and instant way to resolve issues. The hype is there, but how have chatbots performed in practice? And can human-led customer service channels (like call centres and live chat) still offer value in the debt recovery process?
The biggest experiment in chatbots to date is Facebook’s integration of bots into its Messenger app, used by 1.3 billion people. The results are mixed. The Messenger chatbots failed to complete 70% of user requests and only 30% of all interactions were completed without the problem being elevated to a human.
The technology is inhibited because chatbots operate on a decision tree model. They respond to specific keywords that users enter. When the conversation veers from the script, the bot struggles to handle the query.
That’s a major obstacle in sensitive financial situations like debt collections, since customers often require bespoke help with their debts.
The long-term solution to these challenges lies in artificial intelligence (AI), which is increasingly capable of recognising complex requests by mimicking how humans understand language.
But AI isn’t a silver bullet, either. The question of whether a bot should actually imitate a human is a complex one and it raises transparency issues.
In a study of thousands of individuals, Gartner found that transparency is the highest priority for consumers. Trying to pass off a chatbot as a human agent can alienate customers, further complicating the intricate collections process.
But chatbots do offer some benefits. Namely, they aren’t people. They don’t tire and they don’t quit: that’s a handy feature considering call centres have a staff turnover rate of between 30% and 45%.
More generally, chatbots are useful as the first line of customer service, easily providing solutions to common queries. Some bots, like Bank of America’s Erica, can go further, analysing customers’ spending patterns to offer product and financial advice via text messages.
Ultimately, chatbots still have a way to go. So, if they are not the perfect solution, then what other customer service solutions are out there? Let’s look at the two big ones: the familiar call centre and live chat.
Banks – as research by Gartner group shows – are increasingly turning to technology to bring collections into the customer engagement area. Considering this trend, live chat – with its ability to offer real-time convenience to consumers – is an intriguing customer service prospect.
Modern communication is defined by chatting: the number of messaging users is set to top 2 billion this year and will swell by another 500 million over the next three years. It’s a natural communication channel for your customers.
It costs the customer nothing to contact you and it’s cost-efficient for the bank since one support agent can handle multiple cases simultaneously. This isn’t limitless, however; a customer service operation using chat should set a limit per agent, and some chats need to be queued.
And while no customer loves being queued, consumers on live chat, in particular, have high standards. When it comes to customer service, SAP has found 87% of customers on live chat expect a response within 24 hours, and 20% within just one hour.
When you’re dealing with a customer’s complicated and unique debt portfolio, excellent customer service takes time. And it’s worth considering the nature of the questions at play. Live chat works well with online shopping, for instance, as 56% of respondents aged 18 to 34 (and 44% across all ages) prefer live chat to phone.
But live chat doesn’t translate well into the finance space since these numbers change drastically when the questions deal with financial queries, as 79% of consumers across all ages prefer telephone-based customer service.
This brings us to the staple of customer service: the call centre.
IBM estimates that by 2020, 85% of all customer interactions will not be handled by a human agent. But far from making call centres irrelevant, AI, chatbots and self-service technologies will free up call centre employees from routine support requests, so they can focus on more complex tasks.
The debt recovery process is ahead of the curve in this respect. Self-service technologies have lessened the workload of recovery agents and call centres have refocused on technical challenges. Call centres also remain an essential tool for less technologically inclined debtors. After all, your debt portfolio includes a diverse selection of customers.
In the past (especially in the debt recovery process), call centres have suffered due to poor capacity planning. Call centre agents became overburdened and alienated, leading to high staff churn.
Technology has helped manage the workload, though. And the right technology has restored the call centre as the most relevant customer service tool in debt recovery. The right technology is crucial, but often the customer requires a human touch.
Get in Touch