One of the things that leaders say, almost reflexively, is that they are “data-driven decision-makers.” And most organizations, wanting to indicate that they are up-to-date, praise the importance of making data-driven decisions. Consultants and vendors will help with it; big data promises to make it easier and more effective.
I guess all of that is fine, if by making a “data-driven decision” you are saying that you don’t make decisions based on caprice, dart-throwing, or bald self-interest. And truth be told, most decisions are data-driven, in that whoever makes the decision is making it based on evidence, some of which is numerical.
But data-driven decision-making is no panacea, and I worry that the vogue for it hides gaps in wisdom and strategy. Here’s why.
Amassing and analyzing data abstracts it from the actual experience of actual people. In higher education, for instance, the effort to retain students is now augmented by reams of data, telling us the characteristics of students who may be at-risk at our school. But that data hides the fact that we are not trying to retain a cluster of characteristics. Instead, we are trying to retain an actual human being--Jessica, let’s say. There is a lot about Jessica that data can’t tell us, and there is a lot about the data related to Jessica that has nothing to do with retaining her. An effort to retain students (or do anything else for people) that relies solely on data is likely to depersonalize the exact people you are trying to help.
Data may explain why the situation is the way it is, but it rarely tells you what to do about the situation. Colleges routinely come at problems by looking at data. The data may say that nursing students take too long to graduate, or that business students aren’t getting hired at the rate you expected, or that there is a dearth of engineers in your state, for instance. You may not have known those things. But knowing it does not indicate what to do. You could start a networking program for business students. Or you could revise the nursing curriculum. Or you could launch an engineering degree. But none of those steps will necessarily solve the problem, because understanding the shape of the problem does not indicate the shape of the solution.
Finally, putting data in the foreground can hide the fact that there is no strategy behind it. The promise of “data-driven decisions” is that having better data makes it possible to make better decisions. But in practice, having more data and analyzing it more robustly can leave you either seeing more complexity in an issue that was already complex, or with more possible responses to a problem that still has no best answer. So there you are--data in hand. But without a strategy, a worldview into which you can make sense of the data and which will suggest which of the many options is best, you aren’t making a better decision, only a decision surrounded by more complete data.
I am sure that each day, people in my industry make decisions that are better because they have more data. And I am sure that certain types of decisions (about, say, efficiency), are routinely improved by access to more and better data.