Why Have We Stopped Chatting with Chatbots?

A recent survey has revealed that chatbot use is dwindling, with just 8% of customers opting for an AI assistant in their most recent customer service interaction.

ChatGPT Chat concept AI, Artificial Intelligence. businessman using technology smart robot AI, enter command prompt, contact for business information analysis, Futuristic technology
Contact CentreInsights

Published: August 4, 2023

Rhys Fisher

Despite the recent advancements in AI technology, customers appear to be turning their back on chatbots for their customer service needs.

Conducted by Gartner, a survey of almost 500 B2B and B2C customers has revealed that just 1 in 12 respondents used a chatbot during their latest customer service communication.

More worrying still, of the few customers who did use a chatbot, just 25% of them would use one again in the future. Indeed, amongst B2B customers, chatbots were ranked as the least utilised of all customer service channels (7%) – highlighting a clear lack of trust in the AI technology.

But why are customers avoiding chatbots in spite of them being more advanced than ever? And how can businesses address this rising issue?

Crossed wires

The chatbot market is developing rapidly, with 76% of organisations believing that virtual customer assistants/chatbots will be a high-value channel capability for their organisation in the next two years.

Yet, despite customer service and support leaders’ growing focus on chatbots, customer use of them remains low – highlighting a clear disconnect between how businesses and customers view their effectiveness.

The variations of chatbots are confusing customers, making them uncertain about when best to engage.

Part of the issue lies in the all-encompassing nature of the phrase ‘chatbot’, which covers everything from digital interactive voice response systems (IVRS) that can provide basic routing, to complex AIs that can recognise customer intents and connect them with the next best actions.

This variation can often confuse customers and create uncertainty as to when it is best to engage a chatbot to resolve their issue.

This divide and uncertainty is evidenced in the survey and is one of the major factors in the plummeting popularity of chatbots.

The data reveals that chatbots are highly varied in their capacity to resolve different customer intent types, facilitating resolution rates as high as 58% for returns/cancellations, and as low as 17% for billing disputes.

Gartner Graph
Source: Gartner

However, while businesses with a deep understanding of their chatbot capabilities and limitations know what issues are a good match, customers do not.

For example, customers are only 2% more likely to use a chatbot for a return/cancellation than use it for a billing dispute, despite the above graph showing that chatbots are far more successful in helping with returns.

The result is that chatbots are underutilised for issues where they can reliably deliver low-cost resolution, and overutilised for issues which should be handled elsewhere.

Gartner Graph
Source: Gartner

Unfortunately for businesses, outlining the issue is the easy part – now they need to fix it.

Maximising your chatbot potential

I know that any talk of ‘reaching your potential’ conjures up images of serious-faced teachers encouraging underachieving/lazy students to knuckle down for exam season – or at least it does for this writer. But in this instance, it is less about working harder, and more about working smarter.

As the survey results show, when utilised in the correct way, chatbots are a massive asset for customer service teams, but too often they are being accessed for issues beyond their capabilities.

Understanding and defining human value in a customer service interaction is essential.

This is a problem that was discussed in depth in a recent episode of MyCustomer’s podcast series, where Colin Crowley – customer experience advisor at Freshworks – had this to say: “You can have challenges where people try to overuse chatbots. So when there’s a case where something clearly needs to go to a human being because it’s far too complex and requires human empathy and a chatbot really can’t solve the problem.

“So that’s really the importance of understanding and defining human value in conversations – understanding where human beings bring in an important component like an X factor.”

Crowley’s sentiments were echoed by the recommendations in the report, which placed the responsibility of bridging the divide squarely at the feet of the organisations – advising them to guide and educate customers on how best to engage with their chatbots.

They also provided specific examples of how this could be achieved:

  • “On-demand” chatbots that clearly advertise issue types they can effectively resolve. For example, “I can check the status of your account.”
  • Chatbots which surface strategically, where they are likely to encounter customers with appropriate issue types.

The need for businesses to help their customers get the most out of their chatbot experiences was also discussed by Michael Rendelman – Senior Specialist, Research in the Gartner Customer Service & Support Practice – who commented: “Because of the relative novelty and diversity of the technology, customers don’t always know which are the best types of issues to bring to a bot.

“This confusion is likely to be exacerbated further with the introduction of ChatGPT and generative AI. Because customers are unaware of what capabilities to expect from any individual bot, it’s up to leaders to guide and educate them on how best to engage with their chatbots.”

Artificial IntelligenceChatbotsConversational AIGenerative AIMyCustomerVirtual AgentVirtual Assistant

Share This Post