How Governments--and the people who run them--can learn and improve: digital technology has given us tools that make a methodical approach to institutional learning more useful and powerful than ever.

• Author: Mechling, Jerry
• Position: Commentary - Viewpoint essay

In managing themselves and their institutions to improve performance --to execute, not just to decide --executives depend on learning. We aren't good at most things until we've done them for a while. If we're smart, however, we learn. But complex organizations like governments typically don't learn as well, or as methodically, as they should.

Fortunately, this kind of institutional-learning process has been long been studied and improved, and is often expressed as a four-part cycle. A widely used version, made famous by John Boyd of the Air Force, is the "OODA" loop: observe, orient, decide and act. Those who complete OODA cycles more quickly and effectively get better than those who don't. Boyd, a military strategist, applied OODA learning to combat operations--most notably in the 1991 Gulf War--where it has often meant the difference between life and death.

Today, with the increasingly powerful and pervasive data-processing and connectivity at our disposal, the OODA loop is more useful than ever. Here's how it works and how it is being put to new uses:

Observe: Gather the feedback needed to learn. Feedback is data on the results of action. Roughly half of all the data assembled since the beginning of time has been collected in the past two years, with more than 90 percent of it in digital form. Social networking has created an explosion of data and other communications, resulting in applications such as "Where's My Bus?" (which shows you where the bus you're waiting for is and when it will arrive at your stop, based on GPS signals). Much more feedback is available, but we've got to look for it and find ways to capture it.

Orient: Use analysis and expertise to assess relationships between actions and results. Numbers were originally created to measure and count acres and animals for collecting taxes. Now we have "big data," opening far more problems to far-deeper analysis. Google, for example, has found that predicting the path of a flu epidemic is made quicker and more accurate by analyzing billions of Internet search questions rather than by waiting on thousands of sick people to show up at hospitals. Also important is using expertise that couldn't be reached before, as with remote X-ray readings, online student tutoring and other support from a distance. Orientation can benefit from analysis and access to expertise, but we've got to get organized to use them.

Decide: Use engagement and algorithms to...