Is it finally the right time to fully embrace speech analytics?
Speech analytics has become fairly commonplace in contact centres over the years. But recent developments in its capability and analytical potential could mean it will prove pivotal in delivering a better, more coherent, customer experience.
I heard an interesting comment from one of our speakers at our Customer Engagement Management Future Focus event earlier this year which was: “there’s no such thing as big data, it’s just a lot of data”. This comment stuck with me as it’s so very true. Every organisation has a lot of data these days – information gleaned over significant periods of time – customer data, market data, transactional data, employee data, statistical data… the list goes on.
This has always raised the ultimate question – how do we make use of all of this data? How do we answer the questions that stakeholders regular ask of us? Now, it’s not so much a question of making use of all the data, but rather finding the data that’s pertinent and reliable to the question. I’m not going to answer the whole challenge here but, in the context of the contact centre environment, I do believe there is one particular technology that is making significant headway in helping organisations make use of ‘lots of data”.’ That technology is speech analytics.
If you found this article interesting, why not check out 'What’s the future of customer segmentation?', from Alan Merlehan, Capita's head of data and customer insight.
Now, I recognise that speech analytics isn’t new, but having worked with this technology for the past 15 years or so, I can honestly say that the adoption rates are sky rocketing this year, with predictions for further increases over the next few years. I put this down to two factors, the first being the accessibility of the underlying technology – processing power has increased significantly and, in turn, the capabilities of the tools have been enhanced through years of experience. The second factor is more about trust – speech analytics is no longer considered witchcraft; it does what it says on the tin and is significantly easier to prove and validate than ever before.
We only have to look at the traditional approach to demand analysis, which often involves manually creating yet more data at the end of every interaction. The downside to this approach is not only is it time consuming, but the matter of trust. Our frontline has a limited amount of time to decide on the call reason, made more difficult by the fact that, typically, the customer doesn’t just want one particular thing, yet the list relies on the agent selecting one thing! On this basis, manual demand capture tools are very one dimensional (eg, it was a billing call) and very subjective- and the upshot is that we simply don’t trust that data!
Where speech analytics comes into play is that, firstly, it is a machine – it has no sense of subjectivity in the same way a person does, therefore it is totally objective. Secondly, the way modern speech classification techniques works means that each and every interaction will have multiple classifications based on the conversation that was analysed. A call was about billing but, at the same time, the customer wanted to understand their eligibility for upgrade, “oh and by the way I have forgotten my password so I can’t check this myself, could you help with that?”. The conclusion is a method of demand capture that goes from 20-40 percent accurate to 90 percent plus, therefore it’s deemed trustworthy, answering questions that were previously very difficult, or came with an element of poetic license.
I have been fortunate enough to be able to follow this journey, having worked with multiple organisations over the years, watching first-hand how this industry has grown and evolved. Business cases are growing from strength to strength with use cases covering every spectrum of the customer and agent experience; from sales effectiveness, satisfaction analysis, demand and failure analysis, quality increases, digital transformation, automated compliance checking… again, the list goes on. Now is the time to embrace speech analytics, it works incredibly well and provides a true ability to make use of our ‘lots of data’ to answer some serious questions to then help us improve.
For a great example of this, our colleagues at Capita O2 will be presenting their use case on their very own Speech Analytics journey at Engage EMEA Awards in London on the 29th November. Their use cases are incredibly compelling and, more importantly, are repeatable by any organisation that wants focus on their data to make a difference.