- There is no substitute for direct observation
- Proposed changes should always be structured as experiments
- Workers and managers should experiment as frequently as possible
- Managers should coach, not fix
This idea of experimenting caught my attention. So often as I'm designing a solution for a customer, they want to try to get the perfect solution first time out of the box. When they don't know what perfect is, they struggle to make decisions. I can recall one client, after 3 weeks of developing a complex interface, others looked at it and we went through another 3 weeks revising it. This was all in development, the real end users didn't have a chance to try it out. No experimentation at all, just managers trying to guess what was needed. No direct observation.
This is just one example of something I see a lot of. Not being open to experimenting. Smart organizations are figuring it out though. The push for a DevOps approach and employing Microservices is a move in the right direction. With DevOps, we can deploy something and if it doesn't work, we can have a different solution out in weeks or maybe months, not quarters or years. With microservices, we get away from the monolithic applications and have a bunch of small, independent components. If one doesn't work quite right, the whole system doesn't fail.
So how open is your organization to experimenting? Are failures discouraged or recognized as the first step towards success? Are you trying to get everything perfect or do you recognize that everything is a prototype, even if it's in production?