What They Don’t Tell You About Real-Time Speech Analytics

Implementing speech analytics technology into your daily operations or contact center provides enormous value to your business. You can use existing call recordings to gain a wealth of valuable insights. While many companies want you to believe that a real-time speech analytics solution is somehow better than what you get with other solutions, this is not the whole truth.

This post will discuss what they don’t tell you about real-time speech analytics, and why it creates more problems than solutions for many businesses.

One NPR article explored how emergency call hold times surged after 911 call centers implemented a new software that prompts a series of standardized questions to help the dispatcher determine the urgency of a call and how to respond.

Sounds like a great idea, right? It would be, if it didn’t increase call times. Agents at a Portland, OR emergency 911 center say that the extra steps add time to each call, and that time adds up. Portland is not the only city to implement this software and face the same problem. The Fire Chief in Pueblo, CO says the average call times “increased by about a minute and a half” after adopting the same software.

The 911 centers in Portland and Pueblo are just two of the 3,700 agencies using this software, many of which have abandoned the solution due to slower call times. How does this apply to speech analytics? The short answer is that real-time speech analytics functions the same way as the 911 software. And it shares the same issues.

While real-time prompts are meant to assist agents, in reality, they cause the following:

When you add staffing shortages, you have a recipe for disaster. You also have less flexibility with real-time analytics, especially when the script prompts don’t apply to the situation.

While agent scripts have their place in QA processes, prompted scripts can infuriate your customers, and impede the very efficiency you sought to create. Why? Customers are not stupid. They know a scripted response from a natural response. So when a customer notices the script during the call, it creates a negative rapport with the agent. Already the call is off to a poor start.

In fact, according to one study, “69% of customers say their experience is improved when the agent doesn’t sound like they are reading from a script.”

Your typical speech analytics solution without real-time prompts offers reduced handle times because it has the benefit of hindsight. For instance, using silence and overtalk detection, managers can quickly identify calls with long periods of silence. Managers can then identify and fix the problems behind the silence and reduce the Average Handle Time (AHT).

Other speech analytics tools are also far more valuable than prompts. Sentiment analysis identifies calls that have language or acoustics that indicate negative reactions from your customers. You can segment these calls and determine the reasons people are upset so that managers and agents can work together to resolve common customer concerns.

A speech analytics solution also provides sales insights. By monitoring how different groups respond to your sales offers and marketing messages, you can train agents to close deals. Plus, you can avoid missed opportunities since the same call data provides insights that you can use to better target your offers.

The key takeaway here is that while real-time speech analytics solutions claim they increase efficiency, they do just the opposite. If you’re looking for a speech analytics solution to reduce AHT and improve other KPIs, make sure you research the company thoroughly. If they offer real-time analytics as a selling point for increased efficiency, make sure they also offer proof of these claims.

If you’re ready to get some valuable insights into your business and customer base, request a demo with a CallFinder expert to learn more about our speech analytics solution and its many benefits.