How much surveillance is enough surveillance when it comes to development monitoring?
Rant: Background: I work on DORA metrics, DevOps metrics, which basically monitor the development environment across the organisation!
I am currently working on extracting DORA metrics for a client in US. The principal software developer on the client side is obsessively pushing for adopting DORA metrics.
Basically below parameters:
- Deployment Frequency
- Lead Time To Change
- MR Frequency
- Change Failure Rate
- Mean Time to Recovery
Monthly reviews are being pushed for and more reasons for WFH justifications are being worked on. Engineering managers are obsessive for metrics. More surveillance, more surveillance, and more surveillance! This is in-evitable, but this early? And, they say, these aren't accurate, take this with pinch of salt, but think and justify 😂
So, what's the scene in your organizations, is it being the industry norm nowadays?
The problem is that most orgs (and people) don't know how to work with metrics and data so it doesn't matter if a fancy dashboard gives them DORA metrics or GOD metrics or whatever.
They will turn metrics into targets (absolute dogshit thing to do). When metrics become targets, they cease to be good metrics. People start gaming them and true/honest employees suffer because they fell behind the curve.
So more surveillance beyond a certain level will not improve the org at all and will just be useless extra costs paid to software providers/consultants.
Oh wow another toxic abbreviation. Hope it dies down
