Seth Grimes of Alta Plana just published an insightful new text analytics report. In short, the industry is maturing, the ROI story is improving, and customer-related applications are main growth drivers. But solutions have some trouble spots that need attention.
In “Text Analytics 2014: User Perspectives on Solutions and Providers,” Seth does a masterful job of presenting how users are adopting text analytics, how they get benefits, key challenges and much more. Seth draws comparisons with similar studies done in 2009 and 2011, but in most cases the changes are not dramatic.
I’ve been following this market for quite a few years myself. I see text analytics as a key technology to listen to customer feedback from unstructured sources, generated both internally and externally. Here are a few highlights that caught my eye.
As you can see in this chart, Voice of Customer and Customer Experience narrowly topped the list of “primary applications.” Many of the other big applications have a customer slant, too.
According to the report, “social” (a broad term meaning online information external to the enterprise) continues to be the most popular source driving text analytics projects. Blogs and other social media, news articles, online comments and online forms were all key sources.
Like any business application, buyers want a return on their investment. I found this chart a little confusing, because it combines measuring, achieved and planning to measure. However, it suggests that improving customer sat/loyalty, growing revenue and doing so efficiently are why enterprises are investing in text analytics.
Makes complete sense to me. Without text analytics it’s literally impossible to efficiently handle any significant volume of unstructured text, which has been a growing source of feedback in recent years.
Overall user satisfaction results were good, with 74% Satisfied or Completely Satisfied. But there’s plenty of room improvement, especially in ease of use where 29% were dissatisfied. Perhaps more troubling (and somewhat at odds with 74% satisfaction), just 48% of respondents said they were likely to recommend their “most important provider” and more than 1/3 (36%) would actually recommend against. Vendors have some work to do here.
The study was sponsored by many of the leading lights (AlchemyAPI, Digital Reasoning, Lexalytics, Luminoso, RapidMiner, SAS, Teradata, and Textalytics) of the text analytics industry but you won’t find a vendor infomercial in this report. Seth notes that he was “solely responsible for the editorial content of this report, which was not reviewed by the sponsors prior to publication.” Well done.
The full report is available for download. Kudos to Seth for conducting this research and providing the results for all.