Packed into the conversations that make up any given business day is a big-data opportunity: text communications. Trends and insights lurk inside the words we use to share ideas, sell our products and services, and ask questions of our teams and clients.
Think of all that text that's flying back and forth as "dark data"—meaning, it's always around us, but we don't perceive it as the rich information source it is. Instead, we continue to pull our key metrics from, say, ad campaigns and sales reports.
But in these standard text messages, you'll find the kind of actionable information that can reach beyond sales and service. This data can affect the root-level decisions that you, as a business leader, have to make to propel your business forward.
To find out how you can uncover the information that may be hiding in your dark data, I spoke with Mike Stringer, senior director of the Data Innovation Group at Citrix. Stringer is one of the experts working on dark data's business implementations—large and small—and he revealed not only what it is, where it hides and how to get at it, but also what dark data stands to do for companies that have their eye on big-data strategies.
Dark Data 101
Let's start with where dark data lives. "There are many areas where data ends up hidden or underutilized," Stringer says. "Think of it as underground … a natural resource like oil."
Essentially, it lurks in the text that makes up the correspondence and online chatter in your company's internal and external forums. Dark data also runs through the text of your support cases and in your business's emails—both outgoing and in-house. It exists within your product-configuration data and in your log files. It can also be found in sales notes. In essence, what we refer to as dark data ends up archived in all the ways that you and your employees already share—and save—information.
Locating this information and then gathering it together, Stringer's team goes to work with advanced text analytics tools—algorithms and processes based on computer-assisted linguistics research at Stanford University. In a sense, they're refining the raw information extracted from company servers, cloud databases, customer records and the like, and creating something that can fuel new business ideas.
"We go through the text and start identifying what all the important verbs are," Stringer explains. "What action is trying to be taken, based on the text? So, in a support case, we're trying to understand what the customer was trying to do, then we go back and look at what's on either side of the verbs. Then we go through the whole thing again and look for key nouns and ask, What were they trying to perform these actions on?
"We start building a library," Stringer says, "so people can start exploring all this text and classify it."
New Data, New Decisions
With the information that dark data analysis can produce, Stringer suggests business leaders can begin to focus on making new decisions in new ways. They can use the information collected to learn how to:
Analyze more efficiently. In the area of customer-relations management, rather than combing forums and business-to-consumer interactions manually—that is, having a human being review them and categorize instances one at a time (and potentially inaccurately)—dark data does this across tracts of information. And it does so according to consistent parameters.
Discover new things. Letting the data speak for itself, in this way, also opens up the process to new discoveries. "It highlights items we didn't know we didn't know about," Stringer says. "It highlights trends we didn't know were going on."
Analyze the business's history to evaluate results. Once those kinds of cases are identified, dark data analysis further illustrates what the business has historically done in reaction to the events now on its radar. If the reactions have been productive, they can then be further reviewed and codified. Likewise, if they're in need of more serious revision, business leaders now know what's gone wrong in the past and can fix it going forward.
An important step in all of this, Stringer says, is that business owners and IT leaders should leverage the cloud. "Data storage is so ultra-cheap in the cloud," Stringer explains. "You can store all this material in its raw format, and you can also archive it. The ability to store and not throw away data is key, even if you don't know what you're going to do with it yet."
It's a proactive idea, like creating a savings account for future business intelligence. Business owners may well be sitting on their own well of valuable insights. But they can protect that resource now, until they're ready to implement the tools that can mine it.
Read more articles on technology.
Photo: Getty Images