Most Big Data success stories tend to be a bit high-falutin, meaning they always seem to be about some unusual situation or some unusually tech-driven, large company.
Yes, it’s nice to know how Google and Yahoo put Big Data to work, and it’s great that windmill companies don’t have to jump on a boat to figure out if the windmill works … but what about your more traditional organization?
In theory, there are lots of what I’ll call “normal world” applications for Big Data, both for IT, enterprise apps, and the lines of business. However, I seldom see actual case studies or real life examples that demonstrated using Big Data in “normal” situations that might be applicable to the rest of us.
That’s finally changing, as more companies step out of the sandbox and put Big Data to use.
A recent Wall Street Journal article offers several examples of how “normal,” traditional companies are using Big Data in traditional lines of business, such as HR, product development, operations and marketing.
My theory is, it’s taken this long because companies weren’t actually sure how to use Big Data. They had to open their eyes wider to see the problems Hadoop could help with, it seems. Caesars Entertainment Corporation’s experience with Big Data is a good example of this.
The company used Big Data to analyze health claims for its employees and dependents to identify situations where health care costs could have been reduced. For instance, one property was using urgent-care facilities less than others, opting instead for the more expensive ER. So, the company launched an awareness campaign at that property. The end results: Fewer visits to the expensive ER, more visits to urgent care, which meant less spent on health care.
But you’ll notice that while the problems are very much traditional problems, the solutions supported by Big Data are not. Big Data is helping organizations solve old problems in new ways, either by accessing new types of data or by viewing the problem itself differently. For instance, Ford avoided a costly marketing study by analyzing data from auto-enthusiast and other auto-related websites.
To me, that suggests Big Data isn’t merely an extension of what’s already been done, but rather a jump in what organizations can do. The growing and clever use of sensor data is proof of that.
It also makes sense that, while we’re learning more about how Big Data tools can help with “normal world” business problems, we’re also hearing more about the Big Data shortcomings, trip lines and other “Big Data, Big Blunders,” as the WSJ put it in an accompany article.
Not surprisingly, that article doesn’t name names. But it does offer four rules for avoiding Big Data mistakes — veterans of data management will find nothing new in the advice to avoid infighting and avoid investing in data for data’s sake.
That may be the one thing that’s not new about Big Data — new tools, new capabilities, new approaches — but the old advice about how to succeed with data still applies.