Any business with a large customer base appears to be prioritising AI, automation and chats to improve customer service and overall customer experience. But is that what the customer wants?
Christian Banhans, managing director of spusu, believes there is away to bridge the gap between the humanity that customers crave, and the technology businesses are keen to integrate.
Despite businesses dedicating time to technology processes, customers still desire that human touch. Why do you think that is?
Automation technologies can be useful for simple routine interactions that take up agents’ time. However, many chatbots suffer from ‘hallucinations’. They can provide misleading or factually incorrect information due to being trained using biased or insufficient data.
For some models, this can happen up to 27% of the time. Chatbots lack the human ability to confirm any uncertainties with a supervisor, which can result in mistakes going unchecked.
Additionally, AI chatbots lack nuanced emotional understanding. Take for example, a customer that is worried they are a victim of fraud after receiving an unexpected bill. While an AI chatbot would reply with a pre-programmed response, human customer service advisers can empathise with the customer and reassure them while the issue is investigated.
Do you think AI bots will ever get up to the mark that human agents are at?
It’s difficult to say, since most industries have only adopted AI-powered customer service tools recently, so we cannot yet see how the technology will progress over time.
An emerging method of increasing AI accuracy and reliability is known as retrieval augmented generation (RAG), where AI models gather data from a custom database. For customer service chatbots, this could include information from internal documents, such as the company’s knowledge bases, contracts and usage policies.
Although this approach is promising, businesses should be wary of allowing AI models access to sensitive customer data stored in these databases, due to cases where chatbots have leaked confidential information.
Although natural language processing (NLP) methods allow AI to determine whether a statement has a positive, neutral or negative sentiment, it’s not possible for chatbots to express empathy like human customer advisers. Therefore, it’s highly likely that there will be the need for a hybrid approach between bots and agents.
Do you think there is a risk of organisations developing an over-reliance of bots and AI in customer service?
I think that we can already see this happening as some businesses move away from human customer service agents and toward an automation-led approach.
Chatbots are predicted to be the main customer service channel for a quarter of organisations by 2027.
For some customers, using a digital service to solve their issues may be preferable. For example, gen Z and millennials are notably averse to making phone calls, so younger customers may choose to solve their problem via chat services.
However, having a chatbot as the only option can alienate large groups of customers, especially the 10.2 million Brits who lack the digital skills required to navigate online services. Offering a wider range of customer service options ensures inclusion for these communities.
What are the best ways to enhance the human touch in our digitally advancing world for the best customer service?
Offering excellent customer service relies on having a team of experienced agents who are equipped with the relevant technical knowledge, communication and problem-solving skills to be successful. Given ongoing recruitment and staff retention issues within the customer support sector, this may prove difficult for some employers.
However, it is achievable for businesses that make offering high levels of customer service a priority. This means not diverting training and development budgets towards AI tools, as well as providing fair pay and employee benefits and facilitating opportunities for progression.
Any further predictions or pieces of advice for organisations looking to keep the humanity in customer service?
As companies increasingly embrace automation technologies in the handling of customer service issues, it’s important to continuously reflect on customer satisfaction. It’s essential that businesses do not ignore customer preferences in favour of AI’s benefits.
Just under half of respondents in a recent survey say their main reason for using AI is to enhance CX. Yet, between July 2023 and July 2024, which saw increased adoption of AI chatbot technologies, the telecoms sector experienced the biggest fall in customer satisfaction score of all the industries measured by the UK Customer Satisfaction Index.
The main cause of the drop was complaint handling.
While it’s not yet possible to say whether the growing use of chatbots has caused this decline, it’s essential that businesses continue to reflect on their policies.