Skip to main content

Computer says no: How chatbots drive away customers

robot 920x540

Chatbots are a secret weapon in many a company’s marketing arsenal. 

Used by banks, health services, education, travel and government, the AI-driven exchanges offer an instant, on-brand, quick fix for several simple problems 24/7. 

Two out of every three digital customer interactions involve a bot now.

But it’s a fine line between customer satisfaction and instant gratification and letting a robot rub your customers up the wrong way and losing their business altogether. 

So says a new study that aims to pinpoint early warning signs before a chatbot interactions turn into a negative experience.

“For firms, chatbots promise improvements in customer service while enabling big cost savings,” said Dr Ana Canhoto at Brunel Business School.

But when something goes wrong in a chatbot chat, customers tend to blame the company, rather than themselves or their lack of experience, her research finds.

“The negatives experiences identified in our study mean firms end up with unhappy customers, may lose customers or face a PR crisis which makes it hard to attract new ones,” said Dr Canhoto.

Published in The Services Industry Journal, the study shows key chatbot turn-offs are:

  • Authenticity – With many given human names to personalise the experience, customers often customers don’t know if they are messaging a real person or not. People who can’t work it out feel duped, fooled
  • Misunderstandings  – The chatbot asks too many questions, or gives the same answer to different questions
  • Lack of empathy – However clever the technology, when chatbots can’t adapt to their emotional state and register anger, stress or frustration, customers get upset
  • They don’t work – Customers think chatbots can’t solve more complex queries, because their scope is generally pretty narrow
  • They can’t connect – If they decide the chatbot can’t fix their problem, people get annoyed and ask to talk to a human. If they can’t, they end the chat, cut ties with the company or vent their anger on social media, discouraging future customers

Knowing these chatbot fails will help businesses fine-tune both their bots and customer expectations of them to avoid negative experiences, says Dr Canhoto.

“When customer interactions with AI chatbots are negative, it can have serious negative ramifications on service providers, as customers can opt for more costly customer support channels, such as a phone channel.

“In such cases, investment in AI technology intended to result in cost savings might backfire and result in a heavier load on other support channels. Or customers may terminate the service, switch to a competitor, or complain on social media.”