To Do Analytics Well, Focus on the Business IssuesBy Guest Author | Posted 2015-04-29 Email Print
If you want to do well with big data and analytics, you have to look at how and why enterprises have succeeded with BDA programs and how and why they've failed.
Analytics Is Difficult
There’s a final, hidden challenge to the analytics process that is missed by nearly every enterprise, but has profound implications for a right-to-left, problem-first approach to analytics. Doing analytics well is hard. Doing something valuable with those analytics is even harder. Even apparently successful analytics projects don’t always have a positive impact on the organization.
A useful analytics project must be followed by change. Too often, analytics teams are utterly divorced from the teams that might drive that change. Indeed, most enterprises go out of their way to silo analytics, silo their IT staff, and silo their testing and optimization teams. Then they wonder why the results of their analysis don’t produce a real business impact.
By focusing on the business problem first, you have a much easier path to understanding whether the results of an analysis can effectively drive change, which teams and technology will be critical to marshal, and what the process for driving change will actually be. This gives you a vastly improved chance of success.
Like bodybuilders on Muscle Beach, far too many organizations are obsessed with creating a powerful analytics capability, while putting too little thought into what that capability is for. Creating muscles for muscles’ sake is every bit as vapid in the enterprise as it is on the beach.
It’s fun to gawk at a bodybuilder (well, fun for us analytics geeks). However, just as those muscle-bound lifters wouldn’t last five minutes in the boxing ring, many of the most powerful enterprise analytics systems are devoid of purpose and deliver little or no business value.
If you’re going to do analytics well, you need to start by focusing on business problems, and you need to create a real strategy about how you’re going to use analytics. Not a plan about resources or processes or collaboration. Not a plan about the type of technology you need or the number of nodes on your server.
You need to develop a plan that says, “We are going to study problem X, answer question Y and develop model Z because that’s what the business needs.”
Gary Angel is a principal and partner at Ernst & Young’s Advisory Digital Analytics Center of Excellence LLP.
The views expressed herein are those of the author and do not necessarily reflect the views of Ernst & Young LLP.