Finding the Voice of the Customer in Unstructured Data

I was recently involved in some market research pertaining to the convergence of structured and unstructured data. It’s a topic that has been catching interest through the intersection of BI, Search and Text Analytics.

As you know, information channels such as the Web, email, call center narratives, field service notes and survey forms (just to name a few) are creating huge stores of nontraditional data. In fact, studies show that more than 80% of the data within most organizations falls outside of the traditional “structured” data category.

The voice of the customer is not always easy to hear, or more importantly, to understand. There is no doubt though that the words found in unstructured data have life. Nouns with pulse rates … Muscular adjectives … Verbs that dart back and forth across the document. Those words, our language in context, often represents the voice and sentiment of the customer.

And the research showed that organizations are starting to prioritize the use of unstructured data with 23% of those ranked as Best-in-Class rating the topic as a top priority while 58% ranked it as a high priority. The Best-in-Class are also aggressively investigating the future sources from which they will pull unstructured data. Web 2.0 sources such as weblogs, RSS feeds and wiki pages lead the way with 52% currently evaluating or planning to evaluate those sources.

However; the perspective of one of the executives I interviewed is worth noting:

“A lot of what passes for structured information is still just structured data. And structured data is not the same thing as information … it often requires processing to become suitably refined so that BI tools can make sense of it. Maybe we’re not Best-in-Class … but we’re still focused on structured data.”

From your point-of-view (technology power-user, marketing executive, consultant, vendor, academic, etc) how do you see the leverage of unstructured data in order to embrace the voice of the customer?