Page 58 - The Real Work Of Data Science Turning Data Into Information, Better Decisions, And Stronger Organizations by Ron S. Kenett, Thomas C. Redman (z-lib.org)_Neat

What It Means to Be “Data‐driven” 45

• conduct good experiments and research to supplement existing data and address new
questions;
• recognize that the criteria on which they should base a decision can vary with
circumstances;
• realize that making a decision is only the first step; they know they must keep an open mind
and revise decisions if new data suggests a better course of action;
• work to bring new data and new data technologies into their organization;
• learn from mistakes and help others do so; and
• strive to be a role model when it comes to data, working with leaders, peers, and subordi-
nates to help them become data‐driven.

All of these traits are important, and most are self‐evident, although a few require fuller
explanation. First, data‐driven companies and individuals work to drive decision‐making
down to the lowest possible level. This may appear counterintuitive – it seems natural to seek
approval at higher levels. But one executive explained the way he thought about it this way:
“My goal is to make six decisions a year. Of course, that means I have to pick the six most
important things to decide on and that I make sure those who report to me have the data, and
the confidence, they need to make the others.”
Pushing decision‐making down frees up senior time for the most important decisions. And,
just as importantly, a lower‐level person spending hours on an issue is likely to make a better
decision than an executive who spends only a few minutes. Pushing decision‐making down
builds organizational capability and, quite frankly, creates a work environment that is more
fun to work in.
Second, the data‐driven have an innate sense that variation dominates. Even the simplest
process, human response, or most‐controlled situation varies. While they may not use control
charts, the data‐driven know that they have to understand that variation if they are to truly
understand what is going on. One middle manager expressed it this way: “When I took my
first management job, I agonized over results every week. Some weeks we were up slightly,
others down. I tried to take credit for small upturns and agonized over downturns. My boss
kept telling me to stop – I was almost certainly making matters worse. It took a long time for
me to learn that things bounce around. But finally I did.”
Third, the data‐driven place high demands on their data and data sources. They know that
their decisions are no better than the data on which they are based, so they invest in quality
data and cultivate sources they can trust (Redman 2016). Data scientists are well advised to
earn this trust. As a result, they are prepared when a time‐sensitive issue comes up. Further,
high‐quality data makes it easier to understand variation and reduces uncertainty. And finally,
success is measured in execution, and high‐quality data makes it easier for others to follow the
decision‐maker’s logic and align to the decision.
Fourth, as decisions are executed, more data comes in. The data‐driven are constantly
reevaluating, refining their decisions along the way. They are quicker than others to pull the
1
plug when the evidence suggests that a decision is wrong. To be clear, it doesn’t appear that
the data‐driven “turn on a dime”; they know that is not sustainable. Rather, they learn as they
go and modify accordingly.

1 Great decision-makers usually do not explicitly employ Bayes theorem, but many think like Bayesians.