This week we look at our favourite answers for our third #DevOpsFriday5 question - What Do You Think Are The Key Metrics for DevOps? Here's our round-up!
What do you think are the key metrics for DevOps?
“I wrote a blog about this here! I do think though that, even though there is value working towards a standard/generic set of DevOps metrics, every business will have some that are unique and highly important to them and are closely tied to their business goals.” Helen Beal
“I think every single business has to be measured on revenue, costs and profits. Customer satisfaction is another big one. I think the easy answer is how many deploys a day, how quick you develop. But I think that could be a red herring. Different parts of the IT machine move at different speeds and fast for one company is not fast for another. So I would say measuring deployments a day could be a false metric in some cases.” Alan Shimel
“The primary business values delivered by DevOps come from the reduction of waste in the delivery pipeline so the most important metrics have to do with waste reduction for the organizations that DevOps serves. Three key example metrics are the number of times problematic product changes cause the automated DevOps infrastructure to be stopped, the number of times changes have to be rolled back and the up-time of the infrastructure.” Marc Hornbeek
“Deployment frequency, time from initial idea to customer delivery, MMTR, customer satisfaction (internal and external) with delivery of solutions and artifacts. % automation of deployment artefacts (infrastructure as code, application code, automation of regression testing)” Keith Watson
“I will point to “Net Innovation”: the innovation value derived by the organisation MINUS the cost to the organisation in terms of things like maintainability, stability, security exposure, IP exposure, and the like. What we really want is to do is to go fast while maximising NET innovation.” Derek E. Weeks
“Customer satisfaction is always #1 in my book. Too often we focus on IT metrics. What good is it if we get stuff out the door faster if we are not satisfying the customer? Speed does not guarantee that value. Some other important metrics are MTTR, availability, SLAs, and the time it takes from build to deploy.” Mike Kavis
DevOps contributes to acceleration of idea-to-production and the related frequency of feedback cycles as well as to MTTR (mean-time-to-repair). "Repair" should be considered in a broader sense and relates to outages, functional and non-functional defects, usabability deficiencies, etc.. Consider however, that DevOps metrics relate only to a subpath of the complete delivery chain, from the development "done" to "available in production". Michael Schmidt
“A good measure of DevOps 'done' is traceability. Can you point to a release of software running in production and then quickly and easily trace right back to the lines of code changed for that release and the change management tasks, requirements and business drivers ? Can you trace from a requirement to deliver functionality and identify exactly when that was released into production? Another measure of DevOps 'done' is the singular processes. Do you have a standard way to deploy each specific technology that is the same across all environments. Do you have a clear adit trail, a separation of responsibility and a security model that controls who can do what and when?” Mark Roberts
“The key metrics for the business. Isn't that who we are working for? Of course there are other metrics that might give you some clues as to what to tweak next, like cost of IT per business end product, speed of response and implementation. But they aren't the key if you really believe you are part of the company.” Ivor MacFarlane
Want to participate in the next round of #DevOpsFriday5? Do that here.