Big data has endured a rocky ride in recent years. It's climbed to what Gartner describes as the peak of inflated expectations and weathered the subsequent fall into the trough of disillusionment.
But now, Gartner has retired the big data hype cycle. The analyst firm thinks it's time to move the big data discussion past hype and into practice.
John P. Kelly couldn't agree more. He's head of predictive analytics at Berkeley Research Group (BRG), an Emeryville, Calif. strategic advisory and consulting firm.
Kelly claims it's time for CEOs to turn the promise of big data to action — and use it to gain new insights and devise strategy.
Making the Most of Big Data
So how do you prepare your CEO to embrace big data? It's easy, Kelly maintains. Just follow these five strategic steps.
Decide what your company wants to know. Your major components in data science are the data, data science, technology and data customers. As always, it’s smart to start with the customer. What insights does the business want or need? What data products could it create that would satisfy data customers, who are likely different from typical customers? What data has the potential to deliver those insights?
Listen to what your data scientists tell you with regard to data relationships that can be proven. Last comes technology — and the budget associated with acquiring it. "By not leading with the tech-vendor perspective, you and your CEO may be surprised at how small your technology investment needs to be to get what you want out of the data," he said.
Get your data in order. Just about every company overestimates the capabilities of its current state of data, which is often built haphazardly on multiple technology platforms. "Data needs to be gathered, cleaned, structured and integrated for maximum effect," he said. That "takes time, people and money, without visible short-term ROI." But like prepping a wall before you paint it, it's key to a successful outcome.
Embrace analytics. Tim McGuire, head of McKinsey's global Consumer Marketing Analytics Center, predicted "analytics will define the difference between the losers and winners going forward." Kelly concurs: he believes it’s essential to introduce analytics in every board meeting.
"Many board meetings involve a significant level of human bias, but should provide an opportunity to take stock of the company’s situation and recommend adjustments accordingly," he said. "Even better, take it to the next level: go beyond bare spreadsheets and basic analytics and move into the world of predictive analytics, which employs data science to make statistically sound calculations of what the future holds. The board and management can then steer to the future together."
Hire, borrow or rent some real data scientists. If you are building your own data science team, counsel your CEO to have an HR strategy and understanding of the costs to attract and retain the best (assuming that being best in class is the goal of the firm), he said. "Money alone doesn’t do it, and even once you’ve gained some real innovators, retaining their services can be even more challenging," he added.
Go for the gold. As in many fields, you have a spectrum of data science talent ranging from good to great and beyond. Unfortunately, a lot of people just claim to be “data scientists," he said — which isn't that hard to pull off in this age of “point and click” software that allows almost anyone to get into the big data and analytics game. But it takes someone with top data skills to get the best results. "We’d all like a simpler path to unique insights and products from our data, but it’s not often the case. So you need the real deal to be sure of your path. Counsel your CEO to find the best, and don’t let yourselves be fooled by anything less," he said.
Title image by Asa Aarons Smith