Change advisory boards are worse than no change approval process

When you move decision-making away from where work is being done, you’re setting yourself up for a bad time.

The authors of Accelerate observed from having analyzed several years of data that:

[E]xternal approvals were negatively correlated with lead time, deployment frequency, and restore time, and had no correlation with change fail rate. In short, approval by an external body (such as a manager or CAB) simply doesn’t work to increase the stability of production systems, measured by the time to restore service and change fail rate. However, it certainly slows things down. It is, in fact, worse than having no change approval process at all.”

If you set up your software development such that your teams need to genuflect before some architecture board and supplicate to have their voices heard, not only are you encouraging toxic behavior that has a real human toll, but you’re also setting up your software to run a lot worse.

If you’re having quality or delivery speed issues, setting up this kind of board won’t help.

Let’s assume for the sake of argument that the folks who would be on this board know what they’re doing and can actually help your teams. If you genuinely have a skills or alignment gap among your teams, then those folks should be down in the trenches mentoring your teams in the moment that work is done


Like this message? I send out a short email each day to help software development leaders build organizations the deliver value. Join us!


Get the book!

Ready to learn how to build an autonomous, event-sourced microservices-based system? Practical Microservices is the hands-on guidance you've been looking for.

Roll up your sleeves and get ready to build Video Tutorials, the next-gen web-based learning platform. You'll build it as a collection of loosely-coupled autonomous services, developing a message store interface along the way.

When you're done, you'll be ready to contribute to microservices-based projects.

In ebook or in print.