If it looks like text analytics, behaves like text analytics, and is called text analytics, it’s probably text analytics, right? Not necessarily.
A text analytics solution may identify key words and phrases, but that does not ensure any level of comprehension or insight. Text analytics should help tell the customer story and empower your brand to make operational adjustments in an instant.
All technology is not created equal. Take a long hard look at your current text analytics solution and decide for yourself if it’s the real deal.
A generic text analytics solution can be a powerful addition to your Voice of the Customer (VoC) program. A text analytics solution fine-tuned to the nuances of your industry, on the other hand, is invaluable. Many text analytics programs use the same classification model—regardless of industry. As a result, accuracy suffers and customer insights are potentially overlooked. Take our custom-built Monitor™ analytics for example, where we’re able to categorize incoming customer comments in real time, providing your brand with relevant and actionable insights the minute the data comes in.
Customer “moments of truth” are formed instantaneously. Your text analytics solution should be able to keep up with critical functions, which operate in real time, and allow for instant notifications on key issues, questionnaire branching changes, and management reporting. As management sees spikes or changes in customer issues, they can drill down with the touch of a button and view the individual comments fueling a customer experience trend.
Speech-to-text technology allows customers to leave voice comments and have their words transcribed and analyzed in real time. This capability enables management to listen to the emotion conveyed by the customer and opens up additional—and less time-consuming—channels for customers to share their experiences.
The average recall score—the percentage of relevant words or phrases retrieved by a text analytics model—of your standard solution is around 50%. That’s essentially the same odds as flipping a coin. Your chosen text analytics solution should have a recall score that clocks in around 90%. Those are good odds.
Comprehension over Computation
Many text analytics solutions employ a statistical model, which counts words. What they tend to be missing is the use of a linguistic model using a natural language processing (NLP) engine. InMoment’s NLP is powered by IBM’s Watson technology and enables our computers to read customer comments and uncover the customer story. Both solutions have their merits, but a linguistic model excels at uncovering experiential customer data.
The Experience Hub™
Wondering what the Experience Hub is? It’s the platform in which we gather loads of experiential customer information. Some of the most valuable data we collect comes in the form of unstructured customer comments. Because your brand should be able to mine insights from any feedback channel, we’ve embedded our text analytics inside of all our products and services.