The concept of DevOps feedback loops borrows heavily from a number of sources, some of which go back sixty years. One of the first concepts of feedback loops was Plan, Do, Check, ActPlan, Do, Check, Act “also known as PDCA”, “The Deming Wheel” or “The Shewhert Cycle”. “Plan-Do-Check-Act” was first popularized in Japan in 1950 and 1951 during the lectures of a Dr. W. Edwards Deming. The short version of these feedback loops is to plan out your project, try it, examine the results and then incorporate any changes you need to make based on the results.
A newer version of the feedback loop is the Lean Startup concept of build-measure-learn feedback loops. The build-measure-learn process happens to go hand in hand with Agile development since they both have the basic concept of figuring out a problem that needs to be solved and then building a minimally viable product so that we can begin the process of learning as quickly as possible. In Agile we do this through teams building out features in an iterative manner, meaning that during each iteration they create a small portion of the feature until they have built a minimally viable product. As the teams release the feature in an iterative fashion, the entire process of releasing is in a sense the measurement portion where starting with Test Driven Development all the way to the monitoring phase the teams are measuring the feature and learning. A key portion of the measuring and learning happens during an A/B testing phase where teams can receive direct feedback from customers as too their satisfaction with the feature being tested. All along the feedback loop, this method offers a scientific approach, introducing a systematic way for teams to engage in product development.