The Operational Style is the bread-and-butter analytical style of companies today. The Operational Style tells you what happened. If you do not know what happened, it’s hard to set expectations about what will happen.
As I mentioned before, most companies today are dominated by the Operational Style. Managers need reports on financial status, teams need reports on quality measures, HR needs reports on performance to assign bonuses or promotions.
Most operational reports are fairly simple. A matrix of numbers. You have columns and rows. The intersection between the two is a number or a color or an indicator of some aspect of your business. The problem then becomes what to do when you have a lot of columns and rows, a lot of operational data. No problem—our technologist friends then created the cube. Take a column-and-row report and add another dimension, say time. Or by state, by customer segment or by SBU. Then add 10 more dimensions. That’s a lot. Let’s face it, your operational systems collect an enormous amount of data and there are almost limitless ways you can slice and dice the data.
The Operational Style is basic. You have to get this mostly right—not perfect, but good enough to run your business. The key challenge in the Operational Style and the key aspect to enable within this style is to make all of this information more easily accessible and easier to use. Technologists have invented many new software products, some fairly expensive, to help you move through and explore the data. Some managers, for example, just want their weekly operational report as a link in their email (or an Excel spreadsheet). They want weekly bursting. Some managers want the ability to zoom in and out, scan up and down the dimensions, to understand the different relationships between the information. This is the key part of this style.
The Operational Style needs to support simple analytical consumers as well as explorers.
Most people are analytical consumers. Give me the information that I need to do my job. I’ll watch the numbers. Explorers go looking for information and relationships. You need both in your organization and you need to support both using the same data so that the numbers that the analytical consumers use match the numbers that the explorers use. If you do not make the analytics tie out, it’s hard to believe the numbers.
Tie-out does not mean that you need to have all the data in one database and only use one analytical application—that’s actually against agile principles. It does mean that you need to create layers of capability that reinforce tie-out. You need one version of the truth, but you need one version of the truth that is flexible. Most companies take “one version of the truth” to mean something big and consolidated, when in fact it’s a business concept that expresses a need. That need can be satisfied multiple ways, and in some cases, much more cheaply than others.
Next up: The Real-Time Analytical Style.