What is deployment frequency?

Deployment frequency is one of the four critical metrics identified in DORA (DevOps Research and Assessment) research. The deployment frequency DORA framework also tracks lead time, change failure rate, and mean time to recovery. Together, these give teams a clear picture of DevOps deployment frequency and overall delivery performance. Organizations with high deployment frequency can often deploy multiple times per day, while lower-performing teams might deploy monthly or quarterly.

Cycle time and deployment frequency are metrics that measure different aspects of the software delivery process. Cycle time measures the duration it takes for a piece of work to flow through the development process, while deployment frequency measures how often deployments occur. Cycle time begins measuring from the moment work begins until it is deployed to production, with a focus on the speed of delivery for individual work items. Deployment frequency does not focus on the size or scope of each deployment but rather on the rhythm of release.

A team can have a short cycle time but low deployment frequency if they complete work quickly, but combine changes for infrequent releases. Likewise, a team might have high deployment frequency but longer cycle times if they release small changes on a regular basis while larger features take longer to complete.

Lead time measures how long it takes for a single change to go from committed code to running in production. Deployment frequency measures how often deployments happen at all, regardless of how long any one change took to get there.

A team can have a short lead time but still deploy infrequently if they batch several ready changes together before releasing, and a team with a longer lead time can still deploy frequently if they ship small, low-risk changes on a tight, predictable schedule. DORA tracks both because they answer different questions: lead time is about speed per change; deployment frequency vs lead time is really about speed vs. rhythm.

Calculating deployment frequency is a process that involves tracking how often code is successfully deployed to production over a specific time period, such as weekly or monthly. For example, if a team deploys to production 20 times in a 30-day period, their monthly deployment frequency would be 20 deployments per month, or approximately 0.67 deployments per day.

Many teams find it useful to categorize their deployment frequency into bands according to the DORA metrics framework, a widely used deployment frequency benchmark: multiple deployments per day (elite), between once per day and once per week (high), between once per week and once per month (medium), or between once per month and once every six months (low).

When calculating deployment frequency, it is important to count only successful deployments that make it to production, not failed attempts or deployments to testing environments. Additionally, some companies distinguish between different types of deployments, such as feature deployments versus hotfixes, to gain more nuanced insights into their release patterns.

Teams can apply several strategies to improve deployment frequency. These steps involve changes to workflows and cultural changes to facilitate an environment that promotes a good deployment frequency on a regular basis.

• Implement continuous integration and continuous delivery (CI/CD) pipelines.
• Break down work into smaller, more manageable chunks that can be delivered in stages.
• Automation throughout a workflow.
• Standardize environments across development, testing, and production.
• Shift to microservices architecture where appropriate.
• Foster a culture that values small, frequent changes over large batches of work.
• Track and reduce approval bottlenecks in the deployment process.
• Implement monitoring and observability tools to quickly identify issues post-deployment.

Teams need to take time to understand their processes and identify areas for improvement. In this sense, it is also important to consider how metrics are used to measure development team efficiency. They should be understood as signals rather than a spotlight on a specific issue. 

Enji helps teams apply all of the steps mentioned above to their workflows to achieve and maintain a healthy deployment frequency. AI features work in the background to keep teams focused on tasks and moving them through their workflow.

Together, these turn deployment frequency from a lagging number you check monthly into something you can actually act on; a release process you can watch, fix, and repeat.