A few weeks ago, I blogged about a recent McKinsey & Company report on the emergence and impact of “Big Data”. I highlighted the retail areas where significant gains may be achievable by harnessing analytics and big data. In this post, I complete my summary of the report.
The report concludes by identifying the barriers retail executives must surmount to realize the potential of big data.
The first is the mind-set of employees and firms; many people still view IT as a back-office function and therefore as a large cost center rather than as an engine for business growth. In contrast, leading companies in their use of big data understand that their IT initiatives will be a crucial source of competitive advantage. These companies must make sure that business and IT leaders collaborate closely.
Another common obstacle for big data leaders is their legacy IT systems. Many of these systems were installed decades ago, well before today’s big data opportunities were considered or even possible. These legacy systems usually include multiple silos of information generated in incompatible standards and formats so that they cannot be readily integrated, accessed, and analyzed.
Potentially as daunting for retail executives is the task of finding the talent that can execute big data levers. Globally, executives complain about the scarcity of high- quality candidates for these jobs, and many retailers do not have sufficient talent in-house.
The first two obstacles are well-known to business professionals seeking to use IT and data to move the needle for their business. The third obstacle, however, may not be as apparent since it comes to the fore only when the other two obstacles have been substantially addressed.
Executives who start to hire data analysts in droves are in for rude shock. They will discover that it is very difficult to find people who can analyze data and draw actionable conclusions for what the business should do next.
There is no dearth of analysts who can find interesting but useless patterns in the data. They will inundate you with charts, pivot tables etc. and you will have wasted several hours of your time before you realize that there’s nothing actionable in the output. Intelligence isn’t the issue; analysts often have high IQs. What’s missing is the habit of connecting the data to business decisions and relentlessly asking oneself, “So what? How can I improve that decision with this data?”. What “bias to action” is to managers, “bias to actionability”should be to analysts.
Teaching analysts to bridge the “last mile of actionability” is hard but there are things one can do to make it easier. I will share what I have learned over the years in a future blog post.