How a human-AI alliance is reshaping customer service

A new frontier for artificial intelligence is being spotlit among experts: customer service. How can this ever-so-human role find its footing amid the ChatGPT boom?

Since the publication of Alan Turing’s paper ‘Computing, Machinery and Intelligence’ in 1950, the ethics, mechanics and potential fallout that comes with cracking artificial intelligence have been hotly debated.

It’s been nearly 70 years in the making. But in as little as seven months, a new wave of generative AI has swept the globe, prompting both fascination and fear and bringing the reality of its impact on the workforce into sharp focus.

Few industries will be untouched by the AI revolution. Goldman Sachs predicts the technology will boost worldwide GDP by 7% in the next decade, with productivity gains to match. And one sector ripe for disruption appears to be the world of customer service.

“We’re a big believer in it,” says Bogdan Maksak, CEO and CTO of DigitalGenius, which provides AI customer service automation for ecommerce companies. He notes that these developments are already allowing businesses to personalise and industrialise digital communications with customers to an unprecedented extent with astonishing accuracy.

Chatbots have a chequered history for companies looking at customer service, often causing enormous frustrations as customers get caught in a never-ending loop of challenging-to-navigate conversation trees. But generative AI chatbots can actually engage with the questions they’re posed, empathising with customers and better directing them than the tech that has gone before, explains Maksak.

The tech behind chatbots

That’s because generative AI is powered by natural language processing (NLP) and large language models (LLMs), which analyse the text they’re provided, and then use their understanding of what it means and what they know to best respond to customers. 

“It’ll have a huge impact on customer service and automation,” says Irena Cronin, CEO of Infinite Retina, a generative AI research consultancy. “Everyone knows that the automation that currently exists for customer service is pretty bad.”

But that’s changing. Companies are now starting to consider using generative AI through their customer service functions, deploying the technology to harness artificial intelligence’s real-time problem-solving potential.

One ecommerce brand that DigitalGenius works with is using automation for around two-thirds of its tasks. If a customer contacts their business and explains that an order hasn’t arrived, the automated system will help them by sending out a replacement or refund. That company is an outlier in terms of how willingly and wholly it has embraced the idea of automation – most firms that DigitalGenius works with use AI for 20 to 30% of their operations, integrating generative features in some form. And that percentage is increasing as companies’ confidence grows, according to Maksak. He says: “It’s about taking away some of the pains that peak periods have for them.” 

Using AI well

Judicious use of AI within customer service functions could unlock potential cost and time savings. In those peak periods, where, traditionally, businesses would have had to turn to hiring, then training, temporary staff, it’s now possible to outsource the work to AI. 

“One of the things that we see our customers talking about is wanting to turn customer service from what they call a cost centre into a revenue-driving value centre,” says Maksak. Artificial intelligence can help enact that change by automating the tasks involved in customer service queries that don’t need to be handled by humans, freeing them up to work on developing meaningful customer relationships that enable them to turn into something similar to a sales team. 

If AI can help handle queries about billing and payment that customers have, it means that the human customer service agents can devote more time to calling existing and would-be customers who, for instance, have ordered a free sample from an ecommerce company to get their feedback and explore whether they’d be willing or able to move into a full subscription for the product. 

There’s obviously a lot of talk about ChatGPT replacing humans, but it’s about making people’s lives easier and replacing the tasks that aren’t fulfilling them

For Maksak, it’s unlikely that generative AI will ever wholly replace the human in customer service – which is important for those businesses where restive staff may be looking on with worry at the rise of AI. CNBC reports that one in four workers fear they could be easily replaced by artificial intelligence and may see the arrival of any kind of AI as a direct threat to their livelihood. 

Amid the concerns, employers have a responsibility to make their plans for these increasingly intelligent tools clear. For staff, the question they want to be answered is: how will the business incorporate artificial intelligence to unlock their ability to work better, more efficiently and effectively, rather than in place of them? 

It’s important to win onside any naysayers, says Maksak. “There’s obviously a lot of talk about ChatGPT replacing humans, but it’s about making people’s lives easier and replacing the tasks that aren’t fulfilling them,” he explains. Take conversations with customers where a language barrier is at play; AI could act as a real-time translator between the customer and human agent, benefitting both.

The case for humans

AI can provide personalisation at scale in a way that humans can’t and is set to reduce the pile of tasks customer service agents face when they come into work in the morning by at least 20%, by Maksak’s estimation. Beauty Pie has integrated this model with great success. “We’ve been working with DigitalGenius for the last year to improve our CSAT, drive down agent effort, and ultimately remove 30% of queries from hitting our agents,” notes Chandni Bhatt, the brand’s senior manager for member happiness. That said, the highly personal touchpoints that make customer service unique are still very much needed.

“If you’ve got 20% more time in your day, personalised conversations with customers by humans are things you can start to add in,” says Maksak. It benefits employees and also benefits customers by ensuring that when someone gets stuck on an issue that is unique, the staff members involved can help unpick the gridlock, offer personal opinions and take the time to really understand the issue.

Maksak foresees a future in which customers are triaged based on the level and speed of service dependent on their immediate needs. For relatively binary or objective queries – like “Where is my order?” – the AI can answer immediately. But not all queries will be solvable with AI, which is why Maksak always recommends retaining the opportunity for customers to interact with a human somewhere. It’s a bold, brave future – and one that can be helped, not hindered, by generative AI.

For more information, visit