Social media isn’t always cooperative when it comes to extracting business value from its supposedly rich veins, and companies know it.https://o1.qnsr.com/log/p.gif?;n=203;c=204663295;s=11915;x=7936;f=201904081034270;u=j;z=TIMESTAMP;a=20410779;e=iI’m fascinated by the question. While there are certainly success stories, there doesn’t seem to be a clear path to achieving that success. My sense of the situation is that everybody agrees there’s value lurking in that data, but there’s not a lot of confidence in finding it. That makes for a hard business case.
Don DeLoach, CEO of the investigative analytic database company Infobright, wrote a useful read on the question for Information Management yesterday. Obviously, he’s promoting investigative analytics as the answer, but I think he does it very well.
Much of the column draws on Stephen Swoyer’s TDWI ebook, “Investigative Analytics: The New BI Frontier,” which you can download for free. I tend to really like TDWI reports and often recommend them — but again, Infobright also sponsored the ebook.
The main point is that we often talk about analytics as one thing, but it’s really a continuum from traditional analytics to predictive analytics. In between, Swoyer places investigative analytics.
As I read it, the problem is that most of us are acclimated to a very straightforward approach to data. Our “data culture” relies on traditional analytics, which groups data for us in dashboards, reports and scorecards. Thanks to search and — you know — life, we understand predictive analytics and its question-driven approach.
But social media data doesn’t behave itself when we apply these approaches. It begs to be explored in a more open-ended way, the article argues.
What that means is, if we want to extract business value from social media data, we need to take a lesson from kindergarten students and play with the data.
Investigative analytics lets you do that, he contends, by allowing “non-data scientist users to ‘play’ with social media data by asking iterative questions in near real time, regardless of data volume.”
That’s practically heresy in some organizations, but life’s hard sometimes. Children deal with this by playing, and we shouldn’t be so quick to dismiss the advantage in that.
While you’re on the site, technologists will find good fodder in John Myers’ recent blog post, “Is There a Place for NoSQL in BI and Analytics,” which is a great additional read because NoSQL is a an option for dealing with Big Data, including social media data.