Every day during busy lunch hours at a fast food chain (I leave you to select your favourite), there is a very loud process, however more notably there is also smooth absolute operational excellence. But how “can we know”? Well, we can see. We see the delivered output, i.e. number of customers served, the frequency of the service and the waste of time in between (which is generally very close to zero).
How To Measure
While we can see and measure the smoothness of the operation and effectiveness of the processes at your favourite fast food chain, things are not as obvious when measuring the Agile delivery of your solution. We still have access to measure features that are delivered, the waste in between and the frequency of deployments, but it might not be as easy to derive the productivity KPI’s out of these metrics.
And now enters Value-Stream-Mapping. A lean measurement technique to ensure that the software / solution / features we produce offer continuous improvements to the overall Customer Experience. This is an absolute must for businesses who are open to the idea of prototyping and experimentation. This is as you would need to have tangible data to start with, which can then be used to measure the success and effectiveness of your PoC.
KPIs for DevOps and SDLC
A set of Key Performance Indicators (KPIs) can be measured in most situations.
They are:
- How many software defects you see and how frequent they are
- This reflects the stability and agility of your development processes
- Density of your infrastructure
- How much of your infrastructure is being used at a given time
- Cost of your infrastructure:
- The operational costs of development and deployment
- The time it taken between code commits
- How fast you can develop new software features
- Average time taken to resolve issues
- Time takes to feature release:
- The time to release a feature from planning to production.
- Retention of your staff
Collect Data
Be very methodical in automating the collection of this data. Not each and every of these KPIs are needed. You can pick and choose which one would give you the necessary insight into your operations.
Achieving operational excellence is truly a continuous improvement exercise that requires an experimental-driven and data-driven approach.