Traditional Speech Analytics Don’t Go Deep Enough

Speech analytics has become a tech stack staple for many organizations, especially in healthcare. With the ability to quickly surface keywords from thousands of customer calls, it’s easy to see the appeal. The problem is those keywords aren’t actionable. 

For example, while it’s great to know that 20% of your customers used the word “frustrated” in a conversation, it really doesn’t tell you why they were frustrated, what led up to the frustration, and what you need to do about it. 

Speech analytics word clouds can provide a high-level overview of the words your customers are using, but they don’t give you enough context to know what actions to take.

What this tells us is that speech analytics shouldn’t be relied on as the end all be all solution, but rather a helpful tool. 

Speech Analytics Is a Starting Point

What speech analytics can do well is surface the most prevalent themes from a large set of customer calls.

This word cloud summarizes one customer call. From the words displayed you can tell that bills and benefits are all top of mind for the caller and you may deduce that they feel grateful.

“Bills” and “benefits” are topical-based words. They tell us what customers are talking about when they call.

“Grateful” is a sentiment-based word. It can help us understand how the customer is feeling.

This is helpful, but still not actionable. Why is the customer talking about bills and benefits? What are they saying about these topics? What are they feeling or expressing gratitude about?

These questions can only be answered by listening to the literal voice of the customer. But listening is cumbersome. For starters, it’s time-consuming to listen to customer calls. And even if you did find time to listen to a stack of calls, there’s no guarantee the topics you care about would even come up. 

In order to make the most of speech analytics, you also need an effective and efficient way to listen.

Listening at Scale

Speech analytics becomes a valuable tool when it’s used to help:

1. Surface the most prevalent themes
   AND
2. Curate samples of interactions that contain the most important content

This approach allows leaders to listen to audio segments that pertain to the topics they care most about, without having to hunt and peck through a huge telephony platform to find the gold. And this is exactly what we do at Authenticx. 

Authenticx uses speech analytics to surface the most relevant content from literal customer voices so you can listen at scale. We call this Smart Sampling and it enables you to surface the types of customer interactions that matter, based on your strategic priorities. 

Think of it like selecting a movie from your personalized Netflix recommendations versus trying to find something to watch on traditional cable. Smart Sampling gives you the customized and actionable data that traditional speech analytics is lacking.

Authenticx Smart Sampling

• Listen to the content that is most relevant to the business need
• Get the full story around top keywords
• Focus your analysts’ time by having them listen to the most valuable customer interactions
• Hear directly from customers what’s going wrong, so you know exactly what to fix

Traditional Speech Analytics

• See top keywords with no way to listen for those phrases
• Make assumptions around how top keywords are being used
• Assign analysts a random sample of calls
• Use keywords to make an educated guess about problems that need to be addressed

When you shift your attention from viewing keywords to listening to your customers transformational opportunities emerge. You’re able to identify process gaps, target leading indicators of customer attrition, reduce inefficiencies, improve customer satisfaction scores, and so much more. 

Interested in learning more about Authenticx? Watch this two-minute preview or reach out for more details.