Stories of our impact
Voice analytics tool helps companies better understand their customers
Ever finished a phone conversation thinking ‘I wish there was a way to tell what was really on their mind’? Turns out there is. Following a client ask, our people developed BEAT – Behavioural and Emotional Analytics Tool – which is an outcome-based voice analytics platform. But what does that mean, and what does it do?
In plain English, BEAT is a tool that helps our clients understand what their customers are really thinking and feeling. It does this by analysing on a real-time basis phone calls and other interactions with call centre staff, and mapping what it ‘hears’ against quality assurance guidelines. BEAT then applies a risk-related score to each call, based on a range of language and behavioural indicators such as voice pitch and frequency of certain words.
That means you can better understand your customers to deliver consistently improved service, reduce regulatory risk and lower operational costs. It also means the information can be used by compliance teams to ensure that conduct quality standards are met. Let’s see how it was conceived and how it’s helping financial services companies reduce risk for their customers.
The fallout from failure
The timing of this technology couldn’t be better. The global financial crisis of 2008 prompted a regulatory crackdown on financial services (FS) companies across the world, leaving banks, insurers and asset managers facing more (and ever-more-complex) rules – many of which focus on how firms treat their customers.
For the UK, these new rules, regulations and codes mean increased responsibilities and therefore costs, along with a need to repurpose old IT platforms and processes.
So, when a large FS organisation was called on to review nine years’ worth of customer data (that’s 200,000 individual customers, comprising over 250,000 hours of call recordings and 3 million associated documents) to identify if it had broken the rules around treating customers fairly, it turned to us for help.
“It was a chance to shape a new product from its inception. I couldn’t pass up the opportunity to take something new and exciting to our clients."
Man versus machine
There were two ways we could do it: we could put a team in a room to do it manually, which would take too long to complete, or we could turn to technology. And that’s when BEAT was deployed.
We created a team that looked quite different to what you might expect from a firm that’s most famous for employing accountants: computer programmers, audio engineers, data scientists, and world-leading academics who study how people communicate both semantically and cognitively.
The result is a product that’s delivering over 95 per cent accuracy for the client – which outperforms what a team of humans could achieve, in a consistent manner – and doing so in a fraction of the time. What’s more, we’re also testing BEAT in ‘business as usual’ at the client, which means we’re analysing new calls as they come in to check for vulnerability straight away. With an FS provider’s license to sell financial products at risk if they don’t get this right, the impact of BEAT’s support can’t be overstated. In the client’s words: “The voice analytics solution that Deloitte has developed is a fresh approach to identifying high-risk interactions with customers. Its intelligent, machine learning platform allows for prediction of a customer outcome using a combination of language and behavioural indicators.”