The key to efficient and effective data-driven decision making is not data. While data are important and necessary, the real driver of good decision-making is criteria. Even with the best data, the wrong criteria will still lead you to a bad decision.
However, good criteria are not enough. Criteria are only effective when everyone is using the same set.
Several years ago I was asked to help an executive team who was struggling to make a key decision. Each month they’d receive discretionary marketing funds. The executive would ask his team to run a quick analysis to determine where to spend the money. Each month the team would bring back a recommendation. Each month the executive would reject it. Everyone was frustrated. The team was becoming demoralized by the constant rejection of their work. They thought, “Why do we even bother, he always does the opposite.” The executive was equally frustrated. He knew his people were smart and talented. He also appreciated their effort. However, he felt constantly stuck and unable to reach a decision.
A simple exercise uncovered the problem. I asked everyone to write down the criteria (or questions) that they used to make this decision. Not surprisingly, the team was very consistent with one another. They wanted to put the money in under-performing areas (low market share, low customer engagement, low sales person engagement). Their idea was to fix what was broken. The executive, however, wanted to put the money in over-performing areas (high market share, high customer engagement, high sales person engagement). His idea was to put the money where he knew it worked.
Regardless of what data was analyzed and presented, they were never going to agree on the decision. Ironically, they all agreed with the analysis. They agreed on which areas where under-performing and which were over-performing. That wasn’t the problem. The problem was that they disagreed on how those results drove the final decision.
Once discovered, the problem was quickly resolved. We had a quick discussion about what they were trying to accomplish and how they wanted to make the decision. From then on, the team would analyze and present the data using the agreed upon criteria. Suddenly there was a lot more agreement and the decisions took considerably less time.
In a sense, this group was lucky. They were using the same data points to make the decision. The just disagreed on how those data points should be interpreted. More often I see groups where each person is using a different set of data points all together. That’s much worse. Those are the decisions that are never made.
Before collecting, analyzing, and presenting data, it’s important to get alignment on what data is needed and how it will be used to make a decision. Otherwise, you may find yourself either a) collecting data that you don’t need (and wasting a lot of time), or b) having to go back, collect more data, and re-run the analysis (and wasting a lot of time).
Here are five simple techniques for defining criteria and creating consensus:
- Have everyone list the questions that will be used to make the decision. Be sure that you only use “yes” or “no” questions. Compare and refine the lists until everyone agrees. Then, organize the list into three to five high level questions with no more than five sub-questions below each one. (be sure to get agreement on the organization as well).
- Look at the colors or symbols on your reports. The columns that have colors or symbols often reflect the data or criteria that people are using to make the decision.
- Ask someone to walk you through their decision-making process. People can’t always explicitly list their questions. They can usually walk you through a decision. Give them the reports (or whatever data sources they use), point to a specific item and ask, “What action would you take on this and how did you come up with that?”
- Ask them after the fact. Find a decision that was already made and ask them to explain how they came to that conclusion.
- Provide scenarios. Once they make a decision, change the value of one or two of the data points that they used. Then ask if their decision would change. This can help you figure out a) if the data point is even necessary, or b) the thresholds at which the decision changes.
In my experience, 75-80% of inefficiencies in data-driven decision making come from lack of understanding and consensus of the criteria being used to make the decision. Sometimes it happens individually (a person doesn’t think through what they need or how they’ll make a decision) or collectively (each member of the group has different criteria). In either case, problems with criteria create problems with decisions.
Take the time (it’s not that much) to define your criteria before collecting, analyzing, and presenting data. The time spent upfront will greatly pay off in reduced time disagreeing, debating, and re-analyzing data.
—————————
Brad Kolar is an executive consultant, speaker, and author with Avail Advisors. Avail’s Rethinking Data workshop will help you become more effective at defining the criteria that you need to make a decision. For more information, please check out: www.availadvisors.com/rethinkingdata.