How to set up successful speech analytics that benefit your business

If you haven’t tried speech analytics yet and don’t know where to start, this article is for you. I’m going to share my experiences and tops tips for setting up a speech analytics programme. I’ll also show you how to get the most out of speech analytics to support teams across your business.

How to set up a successful speech analytics programme in 4 steps

I work for NewDay, a consumer credit financial services organisation that’s a Sabio customer. I was brought in about three years ago to set up speech analytics for the business. In a recent Sabio webinar, I shared what I’ve learnt along the way. This is my formula for success when it comes to establishing speech analytics in a business.

1. Build the foundations

You need two essential components to create strong foundations for your speech analytics programme. The first is sponsorship from senior managers or a member of the C-suite. At NewDay we had support and sponsorship from our Chief Risk Officer.

The second is that you need to create a speech team. Setting up an effective programme takes time and resources from a dedicated team. It’s not something you can do off the side of your desk for a few hours a week.

2. Set up for success

To be successful, you want to gather useful, informative insights from your speech analytics. To do this you need to build categories, and bring in metadata.

Categories are queries that identify certain conversation types for the calls you receive. In other words, the topics and issues that trigger customer calls are your categories.

Metadata enriches the insights you can get from speech analytics. This includes basic information such as the date and time of a customer call and which agent responded. But you should also bring in customer data from other parts of the company. For example, what was the customer’s account status at the time of their call? Have they recently completed a payment?

All these details help to build the full picture of what’s happening in the contact centre and across the company. Ultimately, you can use this information to better support your customers.

3. Produce focussed, actionable insight

This is the really interesting part! If you’ve followed steps one and two, you can now start to collect huge volumes of data full of juicy insights. However, there’s no point collecting data for data’s sake.

Think about the information you really want and need. Have a plan for how you’re going to use it. Keep the data focussed so you can make the greatest impact.

4. Align the insight to a programme

Once your analysts have gathered and presented all your data, you don’t want the response from the business to be “so what?”. It’s important to show how this information can be actioned by the business. The best way to do that is to establish a programme of work. At NewDay, we created the Customer Issue Resolution Programme, which sits within Customer Services.

We focus our insight on certain areas where we can make improvements. We feed that insight and our recommendations into this programme, which creates a backlog. The customer services team work through the backlog so they become accountable for delivering those improvements.

Our programme has been very successful in using insight to create positive changes across the business. It won us Gold Winner status at the European Contact Centre and Customer Services Awards 2019.

How we use speech analytics

We use speech analytics in a whole host of ways at NewDay, including to:

• Measure the scale of any issues we identify and the impact of fixes we make
• Track and alert us to potentially fraudulent calls
• Monitor the risk of a PCI breach
• Track the volume of previously unheard call types. For example, during the pandemic we received lots of queries about payment freezes. We used this information to influence both training for contact centre staff and company policy changes
• Improve digital literacy. We tracked customer queries and then provided video content to answer common questions. When we introduced a new digital co-ordinator role to support customers with technical issues, we used speech analytics to determine their training
• Redesign our FAQs. Speech analytics identified the frequently asked questions coming into the contact centre, which allowed us to make our FAQs content even more helpful and relevant
• Look at emerging patterns, such as bills not paid on time, to find answers and create new definitions of vulnerable customers
• Track and predict call volumes
• Listen back to relevant sections of calls to hear whether new products or initiatives such as payment deferrals were working.

How you can get the most out of speech analytics

Hopefully I’ve shown you just how useful speech analytics can be. If you’d like to find out more about the different ways we use speech analytics, you can watch the Sabio webinar on demand now.