When ‘Big Data’ Should be ‘Small Data’
“[Companies] don’t know how to manage it, analyze it in ways that enhance their understanding, and then make changes in response to new insights… they don’t magically develop those competencies just because they’ve invested in high-end analytics tools.” –You May Not Need Big Data After All” Harvard Business Review, December 2013
Since the concept of big data became the buzzword du jour, big data has become big business. But a recent study by Harvard Business School suggests that many big-data investments fail to deliver because most companies can’t handle the information they already have. That’s why when it comes to big data, bigger is not always better, particularly for small to midsized companies. Lured by the promise of big payoffs, many companies have sunk millions of dollars into sophisticated data analytics software only to realize they did not have the capabilities to interpret the new insights nor the expertise to turn them into a competitive advantage. For some companies, focusing on small data often makes more sense.
It’s not hard to see why the temptation to jump headfirst into a big-data project can be strong. Giants like Amazon, Google, and Walmart showcase how an entire enterprise can be built around the interpretation of unfathomable masses of data. These companies have perfected the science of gleaning — and capitalizing on — detailed insights about customer behavior. (For example, Walmart was able to pinpoint something as specific as what kind of Pop-Tarts customers stock up on before a storm — strawberry.) With similar analytics tools now available to companies in all kinds of industries, the opportunity to turn hype into hope may be irresistible.
Companies within the logistics and supply chain industries don’t seem to be impervious to the draw of big data. In fact, a survey conducted by Supply Chain Insights found that one fourth of respondents had a big data initiative in place and 65% planned on launching one in the near future. A full seventy-six percent of survey respondents viewed big data as an opportunity. The promise of benefit from the theoretical application of big data no doubt sharpens its appeal. A supply chain company could on the demand side, for example, determine to use big data to map all the quotes and online searches that never became orders and change its marketing strategy based on a newfound understanding of how the purchase of one product leads to the purchase of another. On the supply side, big data could be used to measure the impact of a catastrophic event on suppliers abroad, and consequently, allow the company to plan in advance to mitigate the effects on American consumers. These big data benefit examples could lead to significant advantage for companies with the expertise, structure, and knowledge to collect, analyze, and draw strategy cues from large sets of raw data. Unfortunately, small and mid-sized companies usually aren’t well positioned to do so.
Starting with small data, even if you want to eventually head into big data, is a solid strategy that will produce lasting results. To start, clearly articulate what kinds of data you want to collect and begin running a few simple analytics. Choose from which sources you’ll draw data, because randomly scanning everything between heaven and earth will do you no good. Align your goals with your business objectives and turn your analytics professionals loose on the data. If your company doesn’t have in-house analytics expertise, work to attract the appropriate talent; regardless of whether or not you have a new hire, integration and structuring of analytic personnel positions will be a more significant factor in your success than even your use of the most advanced statistical software program. Finally, spend some time determining how the findings should be presented. You’ll want them to be formatted in an understandable manner and to have a clear application for how they will improve your business.
For those of you working in small to midsized companies, what’s your take on big data? What kind of approach would make a successful small-data initiative?