60% of customers don't want companies to use [rubbish] AI in customer service
Here's an interesting post I saw over at CX Today:
A recent study has revealed that the majority of customers do not want companies to use AI in their customer service offerings.
Conducted by Gartner, the findings are based on a survey of almost 6,000 customers across four continents. The results outline a clear disconnect between companies and customers regarding the use of AI.
I'd recommend reading the post, but I'd like to comment on this one.
I have deep concerns about any organisation deploying AI in customer services when the AI is rubbish or next to incapable of answering even basic queries.
But if the AI is good, if it's effective, if it gets the job done quickly and easily, fantastic. Bring it on.
Unfortunately, it's far too easy for companies to adopt what they think is good technology, only to find out it's next to useless. All this does frustrate everyone - customers, executives, agents and colleagues.
Case in point? One of the UK banks I use for my business banking... they've deployed a useless chat function. Utterly useless. Somewhere, someone in that bank thinks they've done a great job. They'll be reading the weekly report that says "our amazing AI system answered 100% of customer enquiries".
But did it really?
Or do you mean that it said, "hello, how can I help you" and proceeded to say "sorry I don't know" to almost every question and then passed everyone to the human agent?
Or, worse – and this happened to me last week - the chatbot got stuck in a loop and didn't give me the ability to escape-to-the-human-agent. It just kept asking me to select from 4 options, none of which was relevant. I got stuck in its 'answer tree'.
So whilst I think it's great that Gartner is doing such research, I think we can all logically agree that:
- Useless Conversational AI is useless. No consumer wants to be exposed to this.
- Great Conversational AI is invaluable, especially if it's essentially invisible and it gets the job done swiftly, via voice, text or both.
I don't care if I'm talking to a human or a Conversational AI. I care about whether I got my thing fixed or resolved.
As you can see with my recent screenshot walkthroughs of Lloyds Bank, NatWest and HSBC... the big High Street banks are sort of getting there.
From the outside, it looks to me like the underlying technology being used by many of these institutions is last generation, not this generation. Certainly there's next to no straight-through resolution of tasks. I had to do all the work myself because the Conversational AI function (or, chatbot) didn't have the capability.
With all of the strategic advisory I'm doing in this space, I always encourage executives to focus on using best-in-class Conversational AI to free-up their contact centres, but not necessarily because of the cost question.
Yes, cost savings can be absolutely compelling with hundreds of FTE able to be redeployed. But I think this is to miss the opportunity. The real opportunity is being able to dramatically increase the amount of high value conversations your teams are having with customers.
Forget for the moment about trying to automate the mortgage application process by Conversational AI. Yes, that would be nice. But, actually, why can't you do this in person? The base-level answer is we don't have enough people. Because they're too busy answering calls from customers about menial low-value tasks ("I need a replacement debit card.")
Get the chatbot answering voice and text calls. Get it to send out statements. Get it to order new replacement cards. Get it to manage the change-of-address process. Get it to do the grunt work – and free up your teams to be able to instantly respond to the high value conversations.
High value, by the way, doesn't just mean "profit" focused activities such as lending or fee income. Answering queries from the elderly or infirm or those who need additional assistance is also high-value.
Anyway: Good work Gartner. Anything that gets dialogue moving on this topic is very welcome.