Voice Analytics Should Be Equal to Text Analytics

July 15, 2011

Your Voice Analytics Should Be Equal to Your Text Analytics

Text Analytics is a hot topic right now. Companies with successful customer feedback programs are using Text Analytics. At its very basic form, Text Analytics provides keyword search and a word cloud of topic frequencies so businesses know what’s important to their customers. But many Text Analytics engines do much more than that: trending, root-cause analysis, automatic comment categorization, and much more.

What about Text Analytics’ up-and-coming brother, Voice Analytics? Only a handful of VoC (Voice of the Customer, or EFM – Enterprise Feedback Management) vendors even offer Voice Analytics. And most of those Voice Analytics engines only provide the basics: keyword search and word cloud topics. Why? Because Voice Analytics engines don’t transcribe the whole comment, they listen for keywords within the comment predetermined by the user. For example, if a fast-food manager wants to stay on top of his location’s French fry quality, his Voice Analytics will be tuned to flag comments that mention “fries.” Then, the manager has to listen to that comment to find out what it says.

Frankly, that’s pretty weak technology. But that’s the state of current Voice Analytics engines.

The answer is yes, they should have equal importance. Mindshare believes that the best way to utilize Voice Analytics and Text Analytics is to transcribe your audible comments into text and then feed them through your Text Analytics engine. The two methods become equal in the quality and quantity of their results. Valuable, actionable insights are extracted from both. Plus, transcribed comments make for easy referencing and provide retainable data for use over and over again.

Just remember, transcribed audible comments provide potential insights. Transcription is near-worthless if you don’t analyze the comments to find usable information.