How are AI-driven Chatbots Affecting Customer Loyalty?

Many organizations are increasingly relying on chatbots for customer service as a reduced-cost way of dealing with inbound calls. Yet, recent Forrester Analytics data shows that customers are far from thrilled by this. They do not trust chatbots to resolve their service issues and remain sceptical that chatbots can provide a similar level of service as humans.

In January this year, a report from Forrester showed that 54% of US online consumers expect interactions with customer service chatbots to negatively affect their service experience. They complained that chatbots don’t understand their questions fully so cannot deliver useful answers, they do not provide an easy path to human assistance, and if they do get transferred to a human agent, the transition experience is far from easy or satisfying.

It appears, therefore, that chatbots are currently sowing the seeds of discontent with American customers, and the companies deploying them are likely to experience a rude and potentially costly awakening. Although this report was based on US research, the implementation of AI-driven chatbots is also growing exponentially in the UK, making it fair to assume that the same is likely to start happening in the UK too, if it hasn’t done so already.

This is explored in more detail in Creating Customer Loyalty, which explores the point that the low-cost versions of chatbots – the ones that simply respond to strings of words with pre-determined answers (which are the type currently used by most organizations) – are likely to be the ones causing the most frustration and annoyance to customers. This is where customers typically experience unhelpful answers to questions or issues, no matter how they attempt to rephrase them. These are the chatbots that will be causing the most irritation and stress, and therefore are most likely to be damaging customer loyalty.

There are, however, more intelligent systems available. These are ones where all answers generated by the chatbot are checked and modified where necessary by a human agent before being transmitted to the customer. These improved answers are then learned by the chatbot and used for future responses. This way, the library of answers grows and improves over time until the majority of questions and issues are being responded to by the chatbot in exactly the same way as they would by a human agent. Yet these types of machine learning, AI-driven chatbots are more costly to install and operate and are therefore currently less prevalent.

In most instances, the key reasons IT and AI are being introduced are to make things quicker, to reduce the need for people to be involved in some transactions or to enable customers. But many organizations are learning that the more their customers are expected to do for themselves, the more likely they are to expect the cost to be substantially reduced. There is already growing evidence of this from the retail, banking, travel, and hospitality sectors. So although some IT or AI implementations may at first glance appear to reduce costs and improve margins in the short term, they could also result in the need to reduce prices in the mid to long term, with a potentially negative impact on sales and profitability.

A final point to remember regarding chatbots is that the technology is called ‘Artificial Intelligence’. Dictionary definitions of the word ‘Artificial’ usually refer to something that is unnatural, feigned, pretended or false - and that’s precisely what it is. Although AI may eventually get very close to impersonating a genuine human response, it will always be artificial. And just as we have grown to dislike false news, we are also likely to learn to dislike false humans from AI-driven chatbots. So, by automating and depersonalising the connections between customers and your organisation, you could be harming the relationships you have created or hope to create with them, and through that, damage future customer loyalty.