Is this the future calling?: How a Datchet call centre is using AI to stop outsourcing

In a call centre in Datchet, a customer service agent picks up the phone to a disgruntled person on the line – ‘problems with your car financing?’, ‘how can I help you?’

Good customer service can be a tricky task to deliver, but this agent is equipped with tools that perhaps few others have available. For example, on this call, there is another party listening.

Artificial intelligence is monitoring the conversation, scanning for changes in voice tone and language, offering insight into when a potential change of tactics is needed on the phone.

There are no call scripts for the people working at ArvatoConnect, but their silent assistant offers another aid in the armoury to help negotiate and resolve queries.

However, this is just one use of AI for the Ditton Park headquartered company which was praised for its use of tech to help 'preserve' customer service jobs in the UK.

“We did have a choice,” the company’s chief growth officer James Towner said.

“We could do what all of our competitors have done; invest in South Africa and India, or the Philippines, in terms of cheap offshore locations.

“And we decided instead of doing that, we'd invest in the UK and we'd invest in AI because we see that as the future.”

The company has 800 employees across the UK, with around 250 based in Datchet, and has undergone a push to take up the new tech, which was praised in a visit by Windsor MP Jack Rankin last month.

Mr Rankin said the company was 'using AI to strengthen, rather than replace, customer service roles based here in Britain'.

ArvatoConnect operates in the public and private sectors, delivering contracts, including customer services, for Government departments, councils as well as major corporations.

Most of the work in Datchet involves handling car financing for an international automotive company.

ArvatoConnect's ambition, Mr Towner said, was to have AI ‘at the heart’ of its operations.

“If you look at the work that is typically moved offshore, it is typically transactional, and it's all the work that is very easy eventually to automate," he said. 

“Our approach instead is to use AI to remove those simple manual tasks and give our advisors the tasks that actually are much more value adding, much more challenging in some respects, but which allows them to actually use their skills.”

ArvatoConnect makes use of existing AI software like Microsoft’s Co-Pilot, and others that its employees will be familiar with.

Employees are encouraged to suggest how AI can be used to improve operations, with ideas fast-tracked for consideration by the company's board, Mr Towner said.

The measures saw AI introduced to help deliver greater efficiency in a Government contract involved in the procurement of public services – and crucially not at the expense of jobs.

Work is ongoing to implement a digital system to help manage complaints, with an AI looking through extensive contract histories for information to save a customer agent valuable time.

Mr Towner, who worked for more than 30 years in the customer services industry, previously ran customer contact centres at Thames Water and also worked at HSBC.

That experience, he said, had shaped his view of what good customer service should be as well as how technology should, and should not, be deployed.

“It’s all too easy in our industry to automate or use AI on bad processes simply to actually remove the cost,” he said.

“But if you do that, it doesn't take away the fact that it's a bad process still or it's a process that shouldn't exist.”

Mr Towner also insisted the benefit of AI to jobs was evidenced in that ArvatoConnect's motoring finance arm had doubled in size in the three years since he joined the company.

He said that while the company’s own research found most people prefer speaking to a person, that did not mean there was not a role for AI to help in some conversations going forward.

“It doesn't necessarily mean that people with vulnerability don't want to interact with AI,” he added.

For customers in a vulnerable position, such as living with financial hardships, the neutrality of a robot assistant can actually be preferable.

Mr Towner continued: "Our research says if your vulnerability is one of financial vulnerability, then actually quite often - I think it's over 80 per cent of customers - would prefer to talk to an AI than a human simply because of that embarrassment factor.”

He said the challenge - and opportunity - was to strike a balance between AI and human resources. 

“We've embraced it and it's at the heart of our strategy as we move forward,” Mr Towner said. “But I would caveat that by saying it should never be AI for the sake of AI.

“Our mantra - and certainly the way I run things and the company runs things - is that at the heart of it is actually customer service 101.”