We’ve been avoiding writing an article dedicated to DORA metrics for a while (if you’ve been keeping up with the blog and podcast, you know our views on context and metrics). But the truth is, DORA Metrics can help you cultivate continuous improvement among your teams. They can help you boost your engineering velocity without compromising on quality.
The best engineering leaders make strategic decisions backed by data insights. The secret is to know what metrics to track to help you work towards your organizational goals.
So, what are DORA Metrics?
This is your essential introduction to DORA Metrics.
In this post I will:
- Provide a clear definition of DORA Metrics
- Explain why these metrics matter
- Show you how you can effectively track these metrics to make better data-driven decisions for your organization.
What Are DORA Metrics?
Want a DORA Metrics definition?
DORA stands for DevOps Research and Assessment. They’re a Google research group committed to understanding what drives effective and efficient software delivery.
DORA Metrics are four key areas that this research group has found to be reliable indicators of a software engineering teams overall health and performance.
The four DORA Metrics are:
- Cycle Time: How long does it take for code to go into production?
- Deployment Frequency: How often does your organization release to production?
- Change Failure Rate: What percentage of your deployments are unsuccessful?
- Mean Time to Recover (MTTR): When your organization encounters a failure in production, how long does it take for you to recover?
Why Do DORA Metrics Matter?
Tracking DORA Metrics in your organization will help your teams achieve the ideal combination of rapid velocity and outstanding quality in their software engineering. If you want to deliver high-quality software, and deliver it fast, then it pays to track your DORA Metrics.
But as we always say, it’s one thing to track your metrics. Even more important is that you know what to do with your engineering metrics.
Your organization’s DORA Metrics will give you key insights into the areas of your software development pipeline that need to improve. They’ll also help you identify what sort of changes you can make to optimize your processes.
Now let’s look at each DORA Metric in turn and discuss how tracking each one can improve your overall software delivery efficiency.
👉 Sidenote: Athenian will give you full visibility of all four DORA Metrics and help you identify any areas of improvement in your process. With our Align dashboard, teams across your organization can set and track the goals and metrics that matter to them.
The Four DORA Metrics & How To Track Them
⏱️ Cycle Time
Cycle Time = The time between a code change, and this code being run in production. This may also be referred to as Lead Time.
On the Athenian dashboard, your cycle time is broken down into four key areas:
- Work In Progress (WIP): The time it takes an item to proceed from first commit to the last commit on a specific branch.
- Review: The average time spent by your team in a review activity in a specific time frame.
- Merge: The average time between an item’s final approved review and the firs merger activity.
- Release: The average time it takes for an item to progress from the moment it’s merged until it’s released.
Think of cycle time as a super metric.
At a glance, it’ll give you a good overview of your software delivery pipeline’s overall health.
Drill down into this metric and you can identify bottlenecks in your pipeline and possible areas of improvement.
But track your total cycle time and you’ll have a good idea of how your strategies and initiatives are contributing to your continuous and autonomous improvement.
🚀 Deployment Frequency
Deployment Frequency = The number of items you deploy to production.
This is essentially a measure of how quickly your teams deliver software. So in short, it’s a measure of your engineering velocity.
This is another metric that will let you track your organization’s overall efficiency. You want your software development pipeline to be agile.
But an agile workflow can actually be a series of gated waterfalls – every pull request must go through a series of steps, and while some of these steps will be automated, many will be manual. And as your organization scales, your velocity will likely suffer.
Working towards autonomous improvement starts with measuring engineering velocity. So the deployment frequency metric will let you diagnose your organization’s overall health, at a glance. Drill down, and you can improve your agility through identifying and optimizing the bottlenecks in your deployment throughput.
Your goal should be to deploy smaller items, more frequently. Smaller items are easier for your software engineers to troubleshoot to ensure quality and reliability, and a high deployment frequency essentially means you’re delivering value to your customers faster, and more often.
Just bear in mind that the Deployment Frequency metric won’t itself you anything about the quality of your output. DORA Metrics are designed to be considered together, and they work best as prompts for where you should dig down into your data to identify areas of improvement.
💣 Change Failure Rate
Change Failure Rate = The percentage of your deployments that fail. Different organizations have different definitions of failure. But it might mean bugs, rollbacks, incidents, customer complaints etc.
Your change failure rate is essentially a measure of the overall quality of your codebase. The lower your change failure rate, the higher quality code your teams are producing.
As you work towards autonomous improvement, you’ll identify strategies and initiatives to help your teams produce better quality code. This might include making changes to your review processes to encourage collaboration of knowledge-sharing or arranging for training and developing to help individuals and teams level up.
Tracking your change failure rate metric will give you an overview of how effective your autonomous improvement strategies are proving to be.
If you’re investing in a certain initiative – such as improving your review process – but you’re not seeing a corresponding fall in your change failure rate, then you’ll know it’s time to course correct and try something new.
Instead of letting you track the deployment failure rate, Athenian lets you track your deployment success ratio. So Athenian will let you track all the key DORA Metrics, but in this case we prefer to show you how well you’re doing, instead of how much you need to improve!
💪 Mean Time to Recovery (MTTR)
Mean Time to Recovery (MTTR) = The average amount of time it takes your teams to resolve production issues.
MTTR is a measure of your organization’s resilience. All software development teams encounter incidents, failures, and roadblocks in their production. What matters is how quickly and effectively you can resolve them when they do happen.
A low MTTR means your production pipeline will experience minimal downtime. It also indicates that you have the systems in place to detect, diagnose, and fix bugs whenever and wherever they occur, as quickly and effectively as possible.
Again, Athenian can give you full visibility of all the data you need to see that’ll help you improve your MTTR. Crucially, it’ll show you the average time it takes for your teams to acknowledge bugs (MTTA) and the average time it takes your teams to resolve these issues (MTTRepair).
You’ve Got Your DORA Metrics. Now What?
DORA Metrics alone won’t improve your engineering organization. This is why this is the first time we’re writing a dedicated post about them: it can be dangerous to rely on only these metrics to access the success of your engineering org, and understand aread of improvement.
It’s just as important that you know how to interpret these metrics – what are they telling you? How do they work together? What sort of strategic decisions will they empower you to make?
Also, while DORA Metrics provide an excellent oversight of your organization’s overall health, they will never tell you the full story. They can act as great springboards for deep dives into your data. But you have to know where to look, and what to look for.
Athenian will give you full visibility of all four DORA Metrics, along with numerous other key metrics that’ll help you set your engineering organization up for success.
We’ll help you identify any areas of improvement in your process, and with our Align dashboard, teams across your organization can set and track the goals and metrics that matter to them.