Processes are a “recipe” we follow to ensure that something is performed. Usually, this is either supposed to fix something we are bad at. Or it is supposed to guarantee that something really bad doesn’t happen.

It’s worth pointing out that something really complex usually can’t be captured in a process. It’s also worth pointing out that you don’t need a process for something you are really good at.

As I consider development to be a creative task, following a “recipe” is usually annoying. The more detailed and restricting the process is, the more annoying it is.

Processes can be either low level, such as: “Any SQL script run in production must be code reviewed by a DBA”. Or high level such as: “Our hiring process consists of these four steps”.

Processes tend to accumulate over the years. We are more keen to add processes than we are to remove them. Before adding a new process, ask yourself if a process really is the right solution to the problem. Does the cost of doing the process every time outweigh the cost of the occasional error?

With that being said, I think a certain amount of process is necessary. But I don’t think process should be the first tool used when solving something. Improving collaboration, knowledge sharing or automation are usually better alleys to investigate before opting for process.