BY PHILLIP LESLIE, UCLA ANDERSON PROFESSOR OF STRATEGY
What’s the big deal about Big Data?
Lately, everyone seems to be talking about data analytics. But this recent infatuation raises many questions: Is it just a passing fad? Why now? Is it applicable to every industry and company? And what does it mean for today’s business-school students?
The reality is that we’ve now entered a period where companies and managers can differentiate themselves by mastering data analytics. Every M.B.A. should be asking: How prepared am I to engage with Big Data? Those who choose to be dismissive, I would caution, run the risk of being redundant in their skill sets the day they graduate.
To begin, it may be helpful to distinguish among three kinds of data analytics – namely, descriptive, predictive and prescriptive. Descriptive analytics is about harnessing real-time data to provide insight into what is happening in your business, whether it be with customers, employees, suppliers or the like. That may not sound impressive. You might think that it’s been a standard business practice for decades. In reality, however, it has taken huge investments in hardware, software and engineering talent to build this form of analytics. For many companies, this is the ongoing focus of their work in Big Data.
Predictive analytics, as the term would imply, is about predicting something, such as inventory stock-outs, call center volumes, fraudulent transactions, communications equipment failures and patient hospitalizations. The effectiveness of the approach relies on the accuracy of prediction (a high “R-squared,” so to speak). If you can predict call-center volumes, you can optimize staffing levels. Similarly, if you can predict equipment failures, you can perform maintenance or repairs ahead of time.
Prescriptive analytics is about testing causal effects. A manager’s daily task is to constantly search for and implement new approaches, like a new shipping strategy, changes to a customer loyalty program, or improvements in customer service. Prescriptive analytics is about testing whether these changes made any difference to sales, customer acquisition, new product listings, and so on. Managers must be able to distinguish between correlation and causation, as well as statistical and practical significance. This is bread-and-butter work for managers, particularly in companies that value evidence-based management or data-driven decision making (like most technology firms, for instance).
In short, regardless of what form it takes, Big Data is now so pervasive that M.B.A.s should take note. Simply use Google Trends to look for search activity for “Big Data,” and you’ll see just how prevalent the trend is. If you want to remain relevant after graduating, ignore Big Data at your own peril.