Now that more executives and IT leaders understand the basics of Big Data, it’s time to start a different discussion: How do you succeed with Big Data?
I’ve found two pieces that will help you strategically answer that question, from planning stages to what you actually need to put in place to successfully launch your Big Data initiative.
Mind you, this is not about how to do the technology part, but rather, how to make sure your Big Data project actually helps the business.
Your first step to a strategic approach to Big Data is to ask the right questions. A CIO.com article “4 Questions to Ask Before Starting a Big Data Initiative,” suggests you cover these issues before you invest:
That may be, but the business can be a little more hard-nosed about these things. Personally, I think of an ROI calculation as a bit of CYA — it provides the financial justification for the project, yes, but it is also a way to monitor your progress and fine tune to ensure you’re achieving the real goals.
Another reason to do an ROI calculation before, during and after the project: With Big Data, one of the goals is discovery. Since you can’t predict what you’ll discover, there’s a good chance you’ll find additional ROI benefits that you couldn’t at first identify. So, from what experts say, it’s a relatively safe gamble that whatever your initial calculations, you’ll do better when the final results are in. (Although I refuse to be accountable for that statement: Proceed at your own risk.)
CIO.com also recommends you perform a cost-benefit analysis to the project, even though some of the benefits may not be immediately measurable.
There are a lot of great points to consider in this piece, but my favorite is this: When you calculate your ROI, keep in mind that the cost is unlikely to increase too much as the data volume does, since Big Data technologies are usually highly scalable, can run on open-source software and commodity hardware.
The second piece that will help you pursue Big Data with confidence is by Jill Dyché, vice president of thought leadership for DataFlux. Big Data projects that achieve their full potential tend to follow seven steps, which Dyché describes in detail:
So what do you think often happens instead? That’s right, the Big Data team gets stuck on acquiring solutions that deal with the first two steps — collect and store — and forget the rest, Dyché warns.
Another big source of Big Data dilemmas: If IT focuses on the technology at the expense of the business requirement, the business folks become suspicious that this is a “resume-building” project for IT.
“Such an environment of mutual cynicism is the single biggest culprit for why Big Data never transcends the tire-kicking phase,” Dyché writes.
Make sure that doesn’t happen. Start by asking the right questions, involving all the stakeholders, and focusing on the business requirements over the technology hype.
For more on a successful start to Big Data, check out Vish Vishwanathan’s recent IT Business Edge column, “Getting Started with Big Data.”