As customers, in recent years we’ve all grown familiar to being told “calls may be recorded for quality management or training purposes” when we call into a contact centre.
But the reality for businesses is that traditionally only very few calls would have been captured, listened to and used, despite the value that could potentially be locked up in these interactions.
However, in recent years, this has begun to change, thanks to the proliferation of speech analytics.
Speech analytics, also sometimes referred to as audio mining, is a technology that recognises speech and converts it into data. This is then structured so that conversations can be analysed for insights. Typically, a speech analytics solution comprises: a speech engine that converts speech to data; an indexing layer that makes it searchable; a query and search user interface to allow the user to define requirements and carry out searches; reporting applications to present the analytics, often in graphical form; and business applications that are provided by vendors to help users with specific requirements.
The first speech analytics products became available for commercial purposes back in 2002, and subsequent years were characterised by steady double-digit growth, with the technology improving all the while. Now estimated to be a $214 million market, there are signs that interest is beginning to ramp up, as Jim Davies, research director at research and advisory giant Gartner, has witnessed first-hand.
“During a recent strategy day with a client, I was asked to document and report on what clients were asking about, and it blew me away the increase in calls related to speech analytics I had received in 2013,” he notes. “It has gone nuts. In that slice of what I do, about a quarter of calls are now linked to speech analytics, whereas a year or two ago it might only have been a couple of percent.”
So what is driving this sudden spike in interest?
Steve Morrell, founder and principal analyst at ContactBabel, believes that the message regarding speech analytics, may finally be getting through, even though it can seem complicated and there has traditionally been scepticism towards it.
“Most contact centre solutions have a specific, easily-communicated reason for purchase, usually around cost savings. The most popular and widespread solutions, such as IVR, workforce management, CTI and outbound dialling, have all had a clear and quantifiable route to cost savings and improved efficiency. Solution providers comment that cost reduction is often the initial driver for investigating speech analytics, as contact centres realise that there is an alternative to making decisions based on minimal data, and monitoring quality manually and patchily.
“However, speech analytics doesn't stop there, and can be used in many different ways to address various business issues. This is an advantage - it is hugely flexible - but it can also make its message to the market more complicated, and to the cynical, it can seem as though speech analytics is claiming to solve every problem that a contact centre could possibly have. However, depending upon how speech analytics is used, it can certainly assist in cost reduction, agent improvement, business process optimisation, avoidance of litigation and fines, customer satisfaction and loyalty improvements, and increases in revenue.”
With this in mind, let’s further explore how speech analytics has the potential to transform the modern contact centre.
Quality management is a mission-critical function to identify contact centre and enterprise trends and provide insight into how agents are performing. Traditionally a very labour-intensive function, speech analytics is revolutionising quality management.
“Most companies have a very crude agent evaluation process, randomly picking calls and supervisors trying to find a good call to evaluate,” says Davies. “But obviously with speech analytics you can classify all the calls accurately and pre-select calls for supervisors to evaluate and almost start doing the evaluation for the supervisors, adding some intelligence to the whole quality management area.”
Samantha Richardson, solutions consultant at Webhelp UK, also notes how valuable this can be at driving overall standards, particularly in the sales environment.
“In any environment where conversion is the main performance indicator, even the smallest insight can mean big improvements to the results,” she explains. “Equally there will always be those top performers who can build rapport and sell with ease and the majority who whilst will do ok, won’t set the sales boards alight. Speech analytics is an incredibly powerful way to unlock these insights as it allows operational and quality managers alike to understand at a statistically significant level what makes the top performers exceed their peers and build their coaching of the majority around these successful behaviours. By having all of the data on top performers together, they can understand patterns and traits and quantify the impact of even the smallest change to conversion.”
“Speech analytics drives a reduction in operating costs for organisations by providing ‘insight’ from calls and recordings which can be used to improve future interactions with the caller,” says Mark Pritchard, customer experience specialist at Kcom. “By understanding root and branch causes around why interactions happen coupled with automated subject classification, it is possible to highlight process problems that organisations didn’t know they had.”
Davies adds: “Speech analytics enables you to do things like increase first call resolution, reduce call length, that type of thing. You can analyse the audio to check the agent has the right knowledge, whether they have resolved the question or not and did they do and say the right things to drive operational efficiency. If an agent is taking six minutes to deal with an address change call whereas everybody else takes three minutes, there is an efficiency gain by sending that agent on a training course to better learn the address change process within the CRM system, because then he will be doing it in three minutes like everybody else, saving three minutes every call.”
Speech analytics also enables organisations to improve sales, by refining selling techniques and providing flags to agents to ensure they do not miss a sales opportunity in the conversation.
“Another area that benefits by speech analytics is revenue generation – looking for opportunities where the caller is a potential upsell or cross-sell candidate and checking whether the agent is approaching that appropriately in terms of their dialogue, their persuasive selling ability and how they overcome objections. So it can be positioned as a revenue generation tool.”
Richardson agrees. “The other great advantage to using speech analytics in a sales environment is that it allows correlation between call drivers and outcomes. So, by building successful sales ‘categories’ an analyst can begin to understand no only which conversations and products yield the most success, but can also unearth golden nuggets like which contact drivers are linked to particular product sales. Information like this can be strategically useful in maximising an agent’s time to their full sales potential.”
Debt collectors need to balance the need to maximise debt recovered with the need to stay within guidelines. Speech analytics can support this.
“There is a very rigid process involved with collection, so it’s important to ensure that agents are following it correctly,” says Davies. “Organisations have to ensure that agents say and do the right things to, for instance, overcome payment objections – if the customer says “I can’t pay this week”, what is the standard phrase the agent should use to counter that.”
One of the most popular aspects of speech analytics is its ability to monitor the level of the agent's compliance with contact centre scripts and regulations - something that is particularly appealing in heavily regulated industries.
“One of the main reasons businesses in the financial services industry are investing in speech is to make sure their agents are compliant,” explains Davies. “If their staff don’t say the right things on a call, the financial repercussions can be hundreds of thousands of dollars. So using speech analytics as a means to make sure that agents are doing and saying the right things on a call is an important way to ensure there is no come back in the future for them.”
The emergence of real-time speech analytics – which makes assessments during the course of the interaction, rather than post-call – also provides further insurance that agents are compliant, according to Donna Fluss, president of DMG Consulting.
“There are lots of regulations such as the Telephone Consumer Protection Act and the Do Not Call legislation in the US, concerning what you shouldn’t do as a business, particularly in the world of collections and outbound dialling in general,” she explains. “As it relates to collections, there are rules where you have to say certain things within specific time frames or you can be fined, such as who you are and what your purpose is. So if you can notify an agent in real-time you may be able to prevent being open to fines.”
Customer experience management
The technology can also evaluate customer satisfaction with company procedures and policies and can rapidly highlight any performance issues that may be damaging the customer experience.
“Speech analytics not only gives organisations the opportunity to better understand the likes of customer satisfaction, customer churn and customer experience, so that improvements can be driven, but by listening to what the customers are saying and picking up on their emotion, other actions can be driven off of the back of it,” suggests Davies.
“Speech analytics’ use in fraud prevention is a newer but quite an interesting area,” notes Davies. “Organisations are just starting to use speech to recognise the voice fingerprint of customers. So you haven’t got to go through the same verification process because you have a voice recording from a previous interaction and so can quickly analyse the customer’s voice against the past call recording and get a match using speech analytics. It’s an uncommon one, but it will be big down the line.”
Art Schoeller, principal analyst of Forrester, adds that real-time speech analytics can build on this further still. “While it’s not a tool that’s used only for fraud, it does allow you to look for certain spoken patterns between the agents and the customer, so that you can look at the activity on an agent’s desktop and you can identify if it’s a conversation that looks like it could be fraudulent and you can take action.”
Enterprise feedback management
“There are also plenty of things that fall outside of the contact centre,” adds Davies. “It could be campaign feedback, competitive intelligence, pricing sensitivity, product feedback and generally things that marketing and sales and product development all value.”
About Neil Davey
Neil Davey is the managing editor of MyCustomer. An experienced business journalist and editor, Neil has worked on a variety of newspapers, magazines and websites over the past 15 years, including Internet Works, CXO magazine and Business Management. He joined Sift Media in 2007.