I’m hard-pressed to think of a business term that has garnered as much attention over the past year as analytics. A search on Amazon reveals more than 12,000 books on the topic—if we only count hardcovers.
To be sure, the process of making decisions based upon numbers and solid data has become de rigueur in many industries and at many organizations. These days, it’s become downright trendy to drop that term at conferences and in meetings.
You’ll get no argument from me on the benefits of evidence- or fact-based decision making. (It’s a key point ini Too Big to Ignore.) Yet there are limitations. Specifically, analytics or “data” cannot guarantee “the right” business decision—much less a successful outcome. Still, all else being equal, decisions made by consulting analytics and data tend to result in better outcomes than those made by pure gut feel. Hence the legendary Charles Babbage quote: Errors using inadequate data are much less than those using no data at all.
Where’s the context?
But errors persist. The era of Big Data and analytics has not brought perfection and utter certainty with it. Far from it. Even with the insight that analytics typically provides, many people continue to make mistakes. There are many reasons for this, but near the top of the list is looking at analytics in isolation. Consider the following real-world examples:
- Sales increase at a retail company XYZ in December (compared to August). XYZ may be doing something right, but remember that retail sales are notoriously seasonal and have been for a very long time.
- Employee turnover plummets in December and January. Maybe management has improved its practices. It’s essential to remember, though, that employees rarely quit their jobs in December. To boot, many stay until their employers pay out annual bonuses.
It’s critical to benchmark numbers, and not just for seasonality.
Even if an organization theoretically measures a valid KPI, in some instances better and far more valuable ones exist. (Read Moneyball if you don’t believe me.) Because of the enormity of Big Data, it’s not only difficult to find a signal in the noise, but the right signal or signals. This is particularly true when organizations fail to embrace new, more powerful applications specifically designed to derive knowledge and intelligence from large, unstructured datasets. Ditto for simplifying unnecessarily complex systems, something that has never been more critical.
Simply and mechanically looking at a dashboard isn’t enough. Critical thinking is still imperative, especially with analytics. More specifically, employees, groups, departments, and organizations should view analytics within the proper business context and ask if they remain relevant.
Make no mistake: Things change faster than ever today. Even well-trodden KPIs need to continually be revisited and, as Mark Twain so astutely observed a long time ago, statistics can still lie.
What say you?
IBM paid me to write this post, but the opinions in it are mine.