Nationwide branch

‘Copilot, not autopilot’: How Nationwide is boosting award-winning customer service with data and AI

Nationwide, the world’s largest building society serving around 17 million customers and with more than 600 branches in the UK, is expanding its use of generative AI to deliver even better experiences for customers and colleagues.

Nationwide already ranks as the number one high street bank in the 2024 UK Customer Satisfaction Index, but this hasn’t stopped it trying to improve further using Microsoft generative AI.

For example, the way they handle customer queries continues to evolve: by using GPT-4 within Azure OpenAI to help generate letters, average response times have dropped from 45 to 10-15 minutes – an efficiency improvement of around 66%.

By using generative AI to support back-office operations, Nationwide can deal with customer queries faster while colleagues are freed up to focus on serving more customers with potentially more complex issues. 

“Generative AI offers incredible potential to improve how organisations serve customers while improving the employee experience, too,” says Srinath Kanisapakkam, Nationwide’s Chief Data & Analytics Officer.

“I’m delighted by the impact of our first generative AI solution with Microsoft, not to mention the speed with which we’ve been able to implement it.”

Responsible approach

An important element of the collaboration with Microsoft is Nationwide’s commitment to a joint responsible AI framework, which ensures that implementation of AI technologies is ethical and transparent.

The building society has established an AI Centre of Expertise, with the support of IBM Consulting, to oversee and deliver all of its AI development work, and an AI Council to ensure that all AI applications conform to responsible AI principles.

Nationwide logo

Paul Ballard, Nationwide’s Technology Strategy & Enterprise Architecture Director, stresses that Nationwide’s interest in AI is not as a way to reduce headcount, but to help employees “maximise their impact”. All GenAI output is reviewed first, he says – a human must always be in the loop.

“Our mantra is ‘Copilot, not autopilot’,” he concludes.

Firm foundations

To make the most of generative AI’s capabilities, organisations need to ensure they have all their data readily accessible in one place. This is often very difficult, particularly for large, complex financial institutions. Trying to build a coherent picture of a customer’s interaction history and needs in this scenario can be particularly challenging.

Swindon-based Nationwide, which had absorbed around 250 smaller building societies over its long history, has used Microsoft Azure to consolidate and centralise its large data estate. It then applied Azure Databricks, a unified, open data analytics platform, and Teradata’s VantageCloud, to analyse and maintain all this centralised data.  

Now Nationwide employees can access the information they need in one place when they need it and have a much clearer picture of each of their customers.  This also supports their ambition for greater personalisation of services for customers.

Ballard says the successful integration of the company’s data to support customer relationship management will also pave the way for more Azure OpenAI applications in future.

He cites contact centre, credit risk assessment, economic crime monitoring, their virtual assistant ‘Arti’, and CO2 emissions reporting, as some examples.

“Generative AI will enable us to offer more tailored digital propositions, improve service, and optimise decision making, risk management and operational efficiency,” he says.

“Whether it’s in customer service or elsewhere, AI gives us an opportunity to evolve not only what we do, but also how we do it. That, in turn, can help us create more connected and engaging experiences for Nationwide members and colleagues.”