In business and organizational management, facts require data to back them up - otherwise they are just opinions. Data is a tricky entity. To have decision supporting data, you must be selective in what data you collect and disciplined in your analysis. We are all familiar with Mark Twain’s aphorism about lying with statistics.

Fact-based decision making is a disciplined process that requires careful thought before processes and procedures are measured in order to collect the right data and then use unbiased, non-judgmental analysis to tease from the data those facts that are important in making decisions and forecasting the future.

You must have useful data to make informed decisions

Collect data

In order to make informed decisions and assess the impact of your actions, you must collect meaningful data. You collect benchmarking information in the beginning and then you collect the same metrics later as the process is being executed. This comparison allows you to know if you are making a difference and if that difference is in the right direction.

Whether your organization produces a product or provides a service, there are processes throughout that can be measured and improved. You can collect data about many variables, so it is critical that you collect data that is meaningful to the decisions you make. First, select variables that are measurable. Generally that means that the variable value can be reduced to a number. Second, and even more important, the variable must be a meaningful indicator of performance. Collecting the wrong data may do more harm than collecting no data at all.

In very large manufacturing operations, a process called Six Sigma is often used as the core data collection process for continuous improvement. Six Sigma has relatively simple, but data intensive measures, such as defects per million. Six Sigma uses benchmarking and data collection applied directly to processes. DON’T FORGET the processes you seek to measure and improve should include validation from external sources like customer satisfaction or market share.

Six Sigma measures may be difficult to apply to complex systems, low product volumes, or service businesses. Below are some examples of data you can collect. The list is not all inclusive, of course, but it should provoke your thinking.

Don’t hoard data

You collect data to understand how your organization is doing compared to its goals and objectives. There are many reasons you might decide to “keep the data to yourself” such as the power information gives you, your concern about the response to unfavorable data, your disinclination to act on the data – none of these reasons are good for your organization.

Get over it. Hoarding data will result in losing credibility and trust. Continuous improvement requires sharing knowledge that is based on comprehensive and timely data.

Select the most useful data to collect and analyze

If you collect data on many variables over any length of time, you may find yourself in the position of drowning in data and starved for information. As a manager and leader, you need to focus on the key data elements that will most directly affect your decision making. You may know already that the vital few pieces of data include sales performance, or number of complaints or increased market share, or assembly down-time. Or, you may not be sure which of the myriad bits of data really are vital.

You or your staff can use statistical and information visualization techniques to isolate key findings and correlate or cluster data into meaningful pieces. Your selection criteria for the “vital few” should be driven by strategic objectives. (See more below on how to find the vital few).

You do not want to get too far down in the weeds of the data detail. Once you have pulled the data on key indicators, your task is to select the top few, say three to five data points, and then pursue improvement of your organization’s performance on those tasks. This reflects your commitment to continuous improvement.

After identifying the vital few and developing objectives for future performance, then you should work with your staff to develop tactics to reach your objectives. You need to be personally involved in monitoring performance on the vital few measures. And, you need to be sure that your interest is being communicated throughout the organization.

Visually display data

One of the most valuable adjuncts to data analysis is data visualization. These are techniques that represent points of data in relation to one another in order to understand current status and trends more deeply and quickly.

Organizations use data to understand where they are, what is happening, and what may happen. Data helps measure progress and identify problems. Data can be boring. But it does not have to be.

Presenting data using visual representations rather than just text or columns of numbers is less boring and can communicate decision- impacting information faster than strings of numbers or text. Visual representations add a dimension to the data that allows the viewer to quickly grab trends and patterns. And, it is the understanding of trends and patterns that facilitate effective actions.

Presenting data visually lets you represent relationships over time, geography, and among or between groups. Well done visualizations will make your point quickly and almost intuitively to your viewers. And, your viewers will remember the conclusions to be drawn from the data better when they can carry that visual summary around in their heads.

Using visualization also makes patterns apparent that might otherwise be hidden. Here is a set of examples on incidence rates. There are five different data views. Although most of them provide some information, the final one yields a pattern that can drive actions.

Look for trends

Looking for trends is part of data analysis. If you look at your data and see something that is seriously harming your bottom line or an opportunity that needs to be grabbed immediately – you take action.

Trends are more subtle than those data points that you jump on right away. Trends are slight directional changes over time or correlations among different data categories collected simultaneously.

Finding trends and making a proactive response allows you to avoid situations that require emergency, all-out action.

Data-driven trend analysis

One approach to trend analysis is statistical. By applying algorithms to data you can see if your performance reflects normal process variation or if there is a real trend in the direction of the data. In the example above, you want to know if last month’s 930 errors were typical or higher than normal. By looking at your average error rates, you see that 930 errors is pretty close to average, that is 930 is within one standard deviation of the mean – certainly not elevated enough to cause you to do something different.

If you want to practice continuous improvement by lowering your errors per month, you might choose to compare each future month against your benchmark. Or, you can create a trend line and measure its slope over time to see if the direction is the desired one, even if the value differences are not statistically significant. Finally, you can plot just the changes in the average or mean. Any of these techniques will give you a window into trends in your organization’s performance.

Intuitive trend analysis

Another approach to trend analysis is intuitive. Instead of using numeric data, intuitive trend analysis relies on knowledge gained on-the-job or in-the-world. Intuitive trend analysis is a sense of what is normal or a feeling that something is unusual. Intuitive trend analysis shouldn’t be shunned because it seems less scientific than a data-driven approach. Business leaders, famous investors, inventors, researchers, and other professionals often rely on intuition-based trend analysis.

Daniel Isenberg, Senior Lecturer of Business Administration at the Harvard Business School in the Entrepreneurial Management Unit, tells us from his research on intuition that senior managers and leaders use intuition to identify when a problem exists, to synthesize isolated bits of data and experience into an integrated picture – sometimes called an “aha” moment, as a check on the output of “rational” or data driven analysis, and when the need to move quickly forestalls formal data collection and analysis. Intuitive identification of a trend can also provide the impetus to collect confirming data.

In a classic case of intuitive trend analysis, Winn Schwartau, author of the book Information Warfare relays a story that just prior to the first Gulf War, a foreign government could have deduced an imminent military action. “We knew the night before the Gulf War because of all the pizza orders heading to the Pentagon” Schwartau said using as indicators the number of people working late based on cars in the Pentagon parking lot and the number of delivered pizzas. Something unusual was going on…

(More below on how to identify and analyze trends.)

Fix problems with facts, don’t blame

When something is going wrong with a project or a product, it is tempting to find out who is at fault and blame them. There is even a holiday (the first Friday the 13th of the year) called Blame-Someone-Else Day. “The dog ate my homework” kinds of excuses.

As you grew older and more mature, you learned that blaming someone did not change the outcome or improve future performance. And, the person who was singled out for blame was not always friendly or helpful afterward. Now, as a leader and a manager, it may still be tempting on occasion to affix blame when there is a problem. But, just as in childhood, fixing blame does not improve the situation. If anything, spending time and treasure looking for someone to blame will delay solving the problem or fixing the situation. Blaming is passive, detrimental, and discouraging. It is immature. Attaching blame to someone or something external is the opposite of taking responsibility.

The only way to change reality is to get the facts and fix the problem. After you have done what can be done and you are back on track for success, then perhaps there may be some merit to understanding what happened. But even here, you are not seeking to blame someone. Rather, you want to know if there is a systemic reason for a bad outcome – a reason that can be addressed so that there are no identical future problems.

There is a longer-term benefit to fixing problems not blame. You model for your employees the importance of reaching your goals even in adversity, you show them that when mistakes happen they must be addressed, and you do not encourage them to find someone to blame themselves.

Some practical guidance

How to find the vital few

Key indicators are those data points or sets that provide the most useful information with the least amount of time required for collection and analysis. For this exercise, pick a large project or function, brainstorm all the indicators or status, progress, or problems you can think of.

Write up in three columns: Type of data, When to collect, How will you use the data?

Then rank the list and select only the top three.

For each selected piece of data, list three reasons WHY is it one of the vital few.

How to identify and analyze trends

Select a major future event in the news – global warming, health care costs, and retiring boomers – choose an event that represents a potential major change for your organization.

Now, answer these questions:

Is the change a trend or a fad?
Is the trend a function of changes in technology, the economy, environment, or society?
How could this event or condition effect your organization?
How will you know?
What information should you collect?
What will customers want or need in the future based on this trend?
What derivative problems or unintended consequences will need a solution?

Fact-based decision making for great leadership

Great leaders make good decisions based on facts, which are based on solid data. Intuition works too, but a spark of creativity often comes after analyzing the data. And, if there’s a problem, great leaders get the facts and fix it. They don't fix blame, they fix problems.

Maximize your leadership potential today! Dive into the secrets of your brain's productivity and lead with clarity. Subscribe to my newsletter for leadership insights delivered direct to your inbox.

Until next time.

Mike

Great Leaders Make Good Decisions Based on Facts