If there is one catch phrase that is top of mind for nearly every service-oriented business these days, it has to be “customer experience.” For example, even in the utilities industry, Aberdeen Research Groups’ shows that the percentage of energy and utility companies citing the need to improve customer satisfaction — a vital and inseparable component of the customer experience — increased from 18% in 2012 to 35% in 2015. One technology that is proving to be a viable solution to the customer experience challenge is real-time speech analytics (RTSA). Read on to find out why.
Manual Monitoring and Coaching is Reactive and Insufficient
In the earlier days of the contact center, when it was more appropriately referred to as a call center, capturing metrics, interpreting data, and coaching agents was simpler. For starters, the data being captured focused almost exclusively on metrics such as how long a customer waited on hold, how long it took an agent to resolve a problem, and the number of calls an agent fielded each day. Managers could use these metrics to hone in on under-performers by listening in on agents’ calls and providing coaching tips and suggestions via real-time screen pops. As call centers have evolved into contact centers, and email, text, and live chat have become viable communication channels, companies had to rethink their service level standards. For example, what was deemed an acceptable wait time for one channel, such as email (typically requires a 24-hour turnaround) was far different from other channels, such as Web chat (response times of 30 seconds or less are the norm).
Trying to apply yesterday’s agent coaching practices to today’s contact centers leads to several shortcomings. For example, let’s say that “number of interactions” is a key metric a company is using to gauge agent productivity and customer experience. How do you know whether agents are really solving the problem at the end of the call (or email or chat)? What if your customers have to call back multiple times and/or try multiple channels before getting assistance? On the surface, it could look like things have never been better. But, deeper analysis may reveal several dissatisfied customers and much-needed changes. And, most importantly, conventional contact center management processes are mainly reactionary — focusing on poor customer experiences after they have already occurred, hoping to prevent future bad occurrences. Wouldn’t it be nice to be able to resolve problems right away? Or, better yet, detect issues before they turn into full fledged problems? That desire is no longer wishful thinking.
Real-Time Coaching: The New Frontier in Customer Experience
One way that savvy companies are helping their agents deliver more efficient, accurate, and memorable customer experiences is by using Real-Time Speech Analytics (RTSA) technology. Unlike conventional data capture tools, which are limited to working with only structured data (i.e. data residing in a relational database), RTSA allows companies to aggregate unstructured data (e.g. audio) from customer interactions across every communication channel. Not only does this capability lead to powerful insights and reporting after a call, it can even be used to identify and correct problems during a conversation.
To illustrate this point, let’s use an example of a customer service agent working for a major carrier discussing a smartphone plan with a prospective client over the phone. The prospect contacts the the carrier after receiving an email about a special package deal for new customers touting $200 per month for 4 smartphones sharing 10 GB of data in a two year contract. Once connected to the agent, she greets the prospect and confirms the reason for his call. While she is talking, a dialogue box located off to the side of her screen confirms she has successfully completed the introduction (e.g. through a green check mark or smiley face). The prospect mentions the special, but the agent misquotes the price as $250. The RTSA software quickly catches the discrepancy and comes to the rescue in real-time, alerting the agent via a flashing box labeled “pricing error.”
In addition to detecting spoken phrase errors and omitted information (e.g. two-year contract) in real time, the RTSA software includes numerous soft speech evaluators, which look for key elements such as:
- Cross Talking. Is the agent interrupting the customer (or vice versa)?
- Speech Ratio. Who is driving the call – the customer or the agent?
- Stress Levels. Are either the agent or the customer displaying concerning emotions?
- Pause Incidents. Are there long pauses or long speaking passages?
- Volume. Is the agent speaking loud enough to be clearly heard by the customer, or are they too quiet?
- Speaking Rate. Is the agent talking at a moderate tempo?
- Call Quality. Is the connection poor, affecting the overall quality of the call?
(To see a short video demonstrating the call center example shown above, click here).
Just think about the potential customer service benefits — especially for companies that form contractual arrangements over the phone like in this example. Not only does RTSA technology assure the key details of the contract are spelled out to the customer, but the customer’s requirements are also captured and verified. For example, if the customer specifies that he wants a blue-colored phone with 64 GB of memory, the RTSA will flag the agent if she accidentally selects the blue phone with standard memory (e.g. 32 GB).
For years, the mantra within contact centers has been: “measure, improve, measure again, and improve some more.” Real time speech analytics not only gives contact center managers a much wider array of measurements by aggregating structured and unstructured data, it also plays an active role in enabling the other vital parts of the equation, which are “improve and improve some more.”
This new kind of coach isn’t exactly IBM’s Watson, but taking advantage of tools like this will certainly help keep your customer service levels up and definitely increase first call resolution.