It’s National Customer Service Week this week, and one of the key themes being discussed is the use of AI technology and chatbots for serving customers. One big debate for the week is whether chatbots will or will not eliminate any inadvertent bias in how customers are treated.
A new study from AI experts Pegasystems and issued for National Customer Service Week sets out what British customers think.
When questioned about previous human customer service interactions, almost half (49%) of British consumers said they have experienced bias as a result of their individual characteristics, beliefs and/or appearance. By contrast, and despite fears that inherent human bias could be transferred onto modern chatbots, only 8% of respondents feel there is a risk that chatbots will be biased.
The survey revealed other interesting insights, including:
- Consumers don’t consider accuracy of information across all their digital channels to be as important as how quick the response is. Consumers ranked accuracy of information across all of their digital channels (i.e. Facebook, Twitter, email etc.) lower than all other characteristics, such as ‘speed of response’, with only 5% of consumers reporting this characteristic as being most valuable. 8% ranked politeness of the chatbot more highly, with 7% saying that emotional understanding and empathy was more important to them.
- Younger consumers are most likely to have faced human bias. Disappointingly, more young people have felt bias in human customer service (61% of 18-24 year-olds), in contrast to 43% of over 55s, suggesting that prejudice is a growing issue, or one that young people are more aware of. This may explain why the younger demographic believe that chatbots are more unbiased in customer service interactions compared to humans, with 60% of 18-24 year-olds expressing that opinion compared to 37% of over 55s.
- Consumers will accept chatbots using their personal information if it means better service – but they are sceptical as to whether chatbots can achieve this. A majority (54%) said that they expect a high level of tailored, highly personalised recommendations from chatbots, although there were some caveats. When speaking to a chatbot using voice communication, more than a quarter (28%) of respondents expected it to understand them and their preferences well enough to be able to offer bespoke recommendations, but still think a human customer service agent is more likely to understand them and their preferences better. One in ten (11%) are in favour of receiving a personalised service from chatbots, but are still getting used to this new type of technology and are concerned that chatbots know too much about them.
- Most customers don’t think chatbots are going to overthrow humans. Almost half (45%) think chatbots will never entirely replace humans, which is logical, because in the same way that human employees can be assisted by intelligent machines with real-time insights, chatbots and AI will provide a layer of augmented intelligence that will help humans focus on anticipating customers’ needs.
Pegasystems marketing and decision specialist Robin Collyer commented: “The survey results demonstrate that consumers are gradually getting used to interacting with chatbots, however they do still have reservations about engaging with them and their effectiveness. This situation is likely to change over time as people become more accustomed to regularly using them, particularly if they can drastically improve service levels.
“Until now, chatbots used in customer service have been fairly basic, however, the addition of artificial intelligence means that chatbots can now become a key component of businesses’ CRM and customer service strategy. By integrating multiple channels and using customer data to make context-aware decisions, companies can help increase customer confidence in both the use of chatbots and brands themselves. In the long run, their use can help overcome bias and will serve as a useful tool to help humans to provide the highest quality customer service.”