Deployment Frequency
This is one post in a series that outlines some key metrics that best-in-class software vendors tend to focus on when designing the teams, process, and product integrations for distributing their application to customer environments.
Definition
The interval your team is releasing new updates for your customers in production.
Detail
-
Deployment frequency is a velocity metric and an indicator of your teams’ efficiency in delivering value to your customers
-
Releasing frequent updates allows you to continually deliver additional value to your customers. You can see the visualization of delivering more value faster with a more frequent release cadence in the example graph below. This example compares different cadences of releases over a 1 year period where the value of the release increases 5% each month, starting at a base value of 100.
-
Having frequent releases reduces risk as the updates will be smaller, which makes them easier to test and deploy.
-
Releasing on a standard cadence helps with update awareness among your customers and can create a habitual update cycle.
-
According to research conducted by Atlassian, 75% of respondents indicated using deployment frequency as their measurement of DevOps success.
Best in Class
Averaging at least one release per month.
How to Measure
-
Deployment frequency can be measured by tracking how often you release new versions in production.
-
You can also measure the number of features and bug fixes included in each release to help determine the value delivered in each.
How to Improve
-
Release size - if you are releasing on an infrequent cadence, it is likely that your releases have large changes in them. Making smaller incremental changes will result in a better understanding of the changes that were deployed in each release and allow you to identify the cause of any failures more quickly and deploy updates faster.
-
Automation & CI/CD - using automation and CI/CD systems for testing and deploying can help increase your deployment frequency by eliminating the need to manually validate code.
-
Best practices - implementing best practices can allow you to increase your delivery speed.
Related Metrics
- Median age of deployed software - having more frequent deployments with fewer changes, and therefore fewer risks, increases the likelihood that your end customers will adopt and install the releases more quickly and decrease your median age of deployed software.