Demand the (Right) Right Data

One of the responders to my last post asked, "Why are managers so tolerant of poor quality data?" One important reason, it seems to me, is that most managers simply don't know that they can expect better! They've dealt with bad data their entire careers and come to accept that checking and rechecking the "facts," fixing errors, and accommodating the uncertainties that using data one doesn't fully trust are the manager's lot in life. Indeed, senior leaders who accept poor quality data can only be judged complicit in a situation that costs plenty, both up and down the organization chart.

High-quality data are indeed possible. It makes a manager's life (at any level) a lot more fun and productive, and the vast majority of managers, especially those in leadership capacities, should become considerably more demanding!

Consider this vignette. I bought my first new car when I was about 20. The car was well-suited to my "poor college student" lifestyle, and I loved it. It was big enough to take all my stuff to campus (and my friends to the bars), and it got good gas mileage. Even better, I could afford it after a summer on the loading dock!

So I was really annoyed a couple of weeks later when I had to return it to the dealer so some rather minor defects could be corrected. My father counseled patience. "It's a complicated machine, Tom," he said. "You can't expect everything to work perfectly the first time."

My dad was usually right, but the next decade proved him wrong, as Japanese and German manufacturers delivered cars with far fewer defects. It took some time for a critical mass of buyers to change their standards. But soon, taking a new car back to the dealer so minor defects could be corrected was no longer acceptable. I remember how great I felt when, a month after purchasing my next new car, I realized I hadn't taken it back!

Managers, at all levels, should develop similar expectations for their data. Managers — senior leaders especially — need to grow clearer about what they want and become more demanding.

First, they should demand that the data are well-suited to the tasks at hand. Data, like cars, can be incredibly complex. The Facebook prospectus provides a case in point: It cites 483 million "daily active users." One might think that means "the average number of distinct users who visit the Facebook website each day." But it turns out that the Facebook definition also counts those that access it indirectly, via third-party sites.

In and of itself, there is nothing wrong with Facebook's definition. It is almost certainly perfect for some uses and completely inappropriate for others. Mark Twain once noted, "The difference between the right word and the 'almost right word' is like the difference between lightning and a lightning bug." So too with data.

Managers must think through exactly what data they need for each situation. Thus, an executive may need a crisp financial summary to set the budget for a marketing campaign, while the marketing manager requires demographic data to plan it; the analyst requires long time series to understand relevant trends; and the sales team needs a few up-to-the-minute transaction details.

Second, managers should demand that the data they use are correct. They shouldn't have to worry that the financial summary is plagued by errors, that the demographic data are woefully out-of-date, that the time series contains unexplained blips, or that some transaction data are missing.

It's all about trust. Managers are entitled to trust both that they are using the "right data" and that the data they use are "right." When they can — and only when they can — the data are fit for use.

To be clear, there is no "data genie" that will immediately grant their data demands. Managers must follow up. They must make certain they know exactly what the data mean and where they come from. Since they cannot expect perfect data, they must insist that those providing the data fully explain the data's strengths and weaknesses with respect to the tasks at hand. And, perhaps most importantly, they must insist that data improve over time.

All that, of course, is step two. Step one, especially for senior leaders, is to admit they have contributed to the problem. Indeed, their lack of attention has allowed a brutal problem to persist far too long. Or, in the words of the immortal Pogo, "We have met the enemy and it is us."

This entry was posted in Technology. Bookmark the permalink.

Comments are closed.