Big data, big risks
Five ways companies are losing money on big data
Ever since Google used it to track outbreaks of the flu, big data has itself spread virally throughout the IT scene and beyond, collecting more and more hype and legendary status as it grows. But behind all the stories of big data wins, there are some serious risks (mostly financial) to large-scale analytics projects. Here’s five of the biggest worries that big data pros around the world are talking about.
1. Starting big
Just because you’re facing a Mount Everest of data, it doesn’t mean you need to set out to climb it right away. Not only are big projects require a high budget, the right talent and structured workflows, but they often also have larger security implications and legal liabilities, as MicKinsey have noted.
By starting out small, analysts can get a feel for the potential reward and where big data analysis makes most sense. But that also doesn’t mean you don’t need a plan. “The biggest mistake organizations can make with Big Data,” Lorain Lawson writes, “is to start without a business objective.” The worst thing you can do is just follow the technology trends and do it because Google is. As Tim Harford writes “The challenge now is to solve new problems and gain new answers – without making the same old statistical mistakes on a grander scale than ever.”
2. Underestimating the cost
Writing in Wired, Marco Visibelli says that big data project managers need to ask themselves a simple question before getting started: “can your Big Data project work without constant human support? If the answer is “no”, then stop.” Just like any decent agent of The Matrix will tell you, “never send a human to do a machine’s job”. It’s expensive.
That being said, the right personnel also play a crucial role. A successful big data project will require creative (and justifiably expensive) individuals that want to do more than organise and analyse data and want to innovate new ways to make sense of it.
3. Overestimating what it can do
A lot of big promises have been made about what big data analytics can do for us. Open University Professor John Naughton wrote in The Observer that the big data bandwagon might well be taking some companies on a bit of a wild goose chase. “I buy stuff both for myself and my kids on Amazon, for example, which leads the company to conclude that I will be tempted not only by Hugh Trevor-Roper's letters but also by new releases of hot rap artists. This is daft, but does no harm.”
But John Webster, writing in Forbes, says that too much reliance on what we think big data tells us can cause plenty of harm. Retailers use analytics “to develop a sense that [they] know that customer on a more personal level. But what happens to customer loyalty and trust when the analytics engine makes mistakes?” E-commerce enterprises in particular would be wise to take big data learnings with a pinch of salt.
4. Being hip with the latest trends
We can all agree that big data analysis is a tad over-hyped. Every month there’s a new technology on the scene and big data’s title as the most-talked-about field of IT has only just recently been stolen by the Internet of Things (don’t get us started on how over-hyped that is). But the success stories of new technologies you read about at McKinsey, Amazon and Facebook don’t apply to all enterprises. As Paul Barsch blogs, “What worked for them, may not work for you.”
As new tools and databases arrive on the big data scene, they bring both new solutions and new problems with them. In spite of an flourishing and exciting pool of innovation, NoSQL, Hadoop and MongoDB have gradually established themselves as the mainstream of big data and are therefore a safe solution for enterprises.
5. Keeping data under wraps
We can all agree it makes more sense to bring data out into the light of day than let dust gather on it. After all, you never know what your data can tell you until you start to take a look at it. But the real question is: who do you show it to?
Tech consultant Eric D. Brown blogs that opening up data access to those that need it (with the proper rights and security) can be far more useful than keeping it for you NoSQL pro’s eyes only. “Some of the most successful companies using big data that I’ve worked with (and heard about) are ones that have opened up their data to their organization,” Brown writes. By taking more of a community approach to analytics, organisations might learn from outside perspectives and rely on the previous experience of third parties to help solve problems.
Feature image: Herkie