Right-Sizing Big Data for Maximum Impact
By Dave Courbanou
In a recent report, IDC predicts Big Data and associated analytic technology services are set to grow at a sizable 31.7 percent compound annual growth rate, and by 2016, revenues in the Big Data space will hit $23.8 billion. As a result, IDC predicts, this demand in growth and technology will be met with a shortage of skills that will drive many to the cloud.
CIOs should be thinking about using Big Data – particularly because their competition is. The predicted growth rate portends more than a handful of businesses will leverage this technology, which takes it out of the niche and into the mainstream.
But walk, don’t run. It’s important to be careful with Big Data because analytics technology used to create actionable intelligence doesn’t automatically create intelligent data.
Dan McKinley, the principle engineer at Etsy.com, notes in a blog that instantly gaining real-time information through analytical capabilities isn’t always the best way these platforms. Instead, it’s important to right-size the analytical window to produce the necessary intelligence.
This may seem like common sense, but McKinely offers an analogy on why it’s important to set data priorities before installing a solution and following the numbers. In the example, he provides data on two identical coins being flipped at the same time. If we analyze the data in real time (not knowing the coins are identical) we may conclude the coins are different based on their propensity to offer a much different heads-to-tails ratio. But if the experiment is completed and data is parsed after, it will show the 50:50 ratio for both coins. This means an individual following real-time information will have offered up inaccurate information.
“Depending on the change that’s being [measured], making any decision based on a single day of data could be ill-conceived,” writes McKinley. “Even if you think you have plenty of data, it’s not farfetched to imagine that user behavior has its own rhythms.”
What’s the best strategy for CIO? Like many sophisticated modern business solutions, Big Data is best approached through a solution provider or expert advisor. But it’s critical that a CIO works across both the C-level suites and employees alike to uncover the information that should be revealed for maximum business impact. In this way, both the Big Data expert and an organization can create a platform that will not produce immature data and can offer up correct assessments inside the right window of time.
Best of all, a CIO that can successfully pull together a Big Data plan will be able to expand on the platform over time, leveraging analytical information across the business and uncovering new spaces for innovation, cost reduction, potentially even building a wealth of data that is valuable enough for a company to sell.
Like almost everything in the enterprise space, planning is key to a successful deployment. But unlike traditional cloud services and SaaS, an effective Big Data strategy will require a more careful approach both pre and post-installation.