Testing in production is a placeholder for our need to be understood
Logic dictates that when a customer tries something for the first time, we don’t know if our customer will be happy or not. Unhappy customers file complaints, demand support and send feedback. Testing in production is a label that expresses our goal to better understand the customer.
You now have is a common goal, customer value, that is split 1024 to 2048 times to be implemented by 10 to 20 agile teams in 200 to 400 software components. You label your validation effort as different types of testing, even if it is just testing.
You deploy your components in different levels. These levels are normally named after something that resembles the phases in the V-model. You label your validation effort as different types of testing (again), even if it is just testing.
You call your last deployment production, a concept that is familiar from the third industrial revolution from the late 1900s. It is characterized by the spread of automation and digitization through the use of electronics and computers, the invention of the Internet, and the discovery of nuclear energy.
Humans have a natural need to be understood. Testing in a production is an euphemism of saying we know, there is an unknown. We want to impersonate the customer, to prevent the exponential cost that is associated with their unhappiness.
Testing is a series of controlled experiments and uncontrolled events.
The author wishes to help the development environment hold the weather and reduce the number of disappointments experienced by millions of people in everyday life.
Originally published at https://www.linkedin.com.
