3 Speech Analytics Myths You Should Stop Believing

The number of misconceptions about speech analytics (SA) are surprising, especially considering that even people who currently use SA technology believe the myths. While the learning curve for SA is steep, understanding that the following notions are false will be a big help. When you know how SA does and does not work, you’ll be able to better utilize to its potential.

Myth #1: It’s Not Necessary to Listen to Actual Calls

It’s not possible to simply guess at the voice of the customer and you won’t be able to set useful SA settings if you don’t first listen to real recorded calls. The only way to figure out which business intelligence will be most useful to your company is to listen to a random selection of recordings from a specific queue. As you dissect recordings, you should do the following:

  • Map call types for specific objectives.
  • Listen to how things are said, not just what is said.
  • Create a ranking of keyword categories.
  • Identify issues that could cause trouble with the software (transfers, hold music, pre-recorded messages).
  • Interpret dialects, accents and colloquialisms.

While listening to numerous calls from beginning to end is tedious, it’s the best way to figure out which words and phrases are most meaningful when it comes to customer intent and outcomes.

Myth #2: Speech Analytics Are a Set-and-Forget Solution

SA make it easy to locate keywords and phrases, but that ability alone isn’t going to give you valuable insight. SA software has to be optimized in order to provide useful and actionable intelligence, taking into account things like the relationship between content and context or specialized language for certain industries. Furthermore, the data you glean from SA won’t be helpful if it isn’t accurate. In order to make sure that the software is at its optimal level of accuracy detection, the speech engine has to be calibrated, tested and tuned until perfect. To do this, search results have to be audited several times over until the speech engine is in harmony with your business objectives. The setup and interpretation phases may be time-consuming and even challenging, but they’re necessary.

Myth #3: Anybody Can Manage Speech Analytics

SA may run automatically, but the software has to be managed on a regular basis, too. There are a range of tasks that must be completed in order to keep SA performing well. While you may have a staff member who’s qualified to take on this role, it’s important to know what’s expected of the person in charge:

  • Software setup
  • Server and database management
  • Compilation, analysis and delivery of key findings
  • Content review

One or more people can take on this role. Since the various steps call for different skill sets, some companies opt to split up the role into three parts

  1. Administrator for setup and management.
  2. Business Analyst for choosing and delivering key findings.
  3. Interactions Monitoring Analyst to review content.

Whoever is in charge of SA has to analyze and communicate the findings, explaining why and how changes should be made.

In order to get the most out of SA, it’s important to control your expectations. SA software doesn’t rely on omniscience. Instead, it’s a smart, calibrated filter that shows you the parts of calls that relate to your specific business issues. SA software provides the tools you need to uncover actionable business intelligence. Without it, it would be nearly impossible to find these subtle points of a conversation. With it, you can apply sophisticated strategies to work toward your business objective.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s