What Elite Performing Teams Measure | An Introduction to DORA Metrics

Being successful in the digital economy relies on being able to answer “how are we doing and where can we improve?” It’s about a relentless dedication to continuous improvement in every aspect of your business. While business and sales data are relatively easy metrics to define and pull, what about measures related to how you’re doing from a development and delivery standpoint? 

Let us introduce you to your new friend, DORA. 

What Is DORA?

DORA (DevOps Research and Assessment) is a research entity (acquired by a little company called Google back in 2018) that has collected and assessed data from more than 36,000 professionals worldwide since 2014. 

Through their annual State of DevOps reports, DORA has identified and validated key metrics that provide quantitative benchmarks for software delivery performance. The teams that excel within all of these measures are the ‘elite’ performers driving exceptional performance within their organization. 

According to DORA’s research, there are four software delivery performance metrics you should measure. Metrics include:

Speed to Deliver Metrics:

  1. Change Lead Time: How long it takes to go from code committed to when that change is successfully running in production.
  2. Deployment Frequency: How often a team deploys code to production. 

Quality Metrics:

  1. Change Failure Rate: The percent of production changes that result in service incidents that require remediation.
  2. Failed Deployment Recovery Time: How long it generally takes to restore service when a service incident occurs.

Historically, we’ve been conditioned to think that if we go fast we forsake quality and if we want quality, we need to go slow. But what DORA has validated is that these aren’t mutually exclusive. You can go fast AND have high quality. And typically, we find that when you’re challenged to improve in one of these areas, you tend to improve in other areas as well.

Ok, So What Are The Challenges?

As with anything, there are always going to be some challenges. 

For one, collecting these metrics can be challenging. With Deployment Frequency, for example, you’ll need to standardize your unit of measurement across teams. Change Failure Rate is a lagging indicator, and often, you don’t know if the change you just made is the actual reason for the failure. And with Failed Deployment Recovery Time, it might be impossible to determine when the problem was actually introduced into the system. Still, don’t let that discourage you. (We said challenging, not impossible!) Trying something is better than trying nothing, and you can always modify as you go.

The hardest part is just getting started. Opt for bias toward action.

Another challenge we see teams encounter is forgetting the purpose behind these metrics. Infrequently, but still enough to warrant a note, we’ve seen leadership teams use these metrics to compare teams, which goes against the continuous improvement spirit of these metrics. These metrics are merely points of feedback that help us answer what can we learn and where can we improve. 

“When a measure becomes a target, it ceases to be a good measure.”

— Goodhart’s Law

What’s the ROI? 

The impact of using DevOps and modern technical practices within teams is huge. According to the 2021 State of DevOps report, when comparing elite performers against low performers, the elites:

  • Deploy code 973 times more frequently.
  • Deploy code 6,750 times faster.
  • Recover from incidents 6,750 times faster.
  • Have a change failure rate that is 3 times lower.

Each of those actions are not only driving real change within IT, they’re enabling the entire company to create better products and serve their customers. 

The State of DevOps Reports show us that each year the elites are becoming more elite. They’re learning faster, delivering value faster and widening the gap between themselves and everyone else. Optimized IT is table stakes for business today, and it’s why it’s so important you start doing something. Because focusing on delivery performance, ultimately, means organizational performance too. 

Breakdown of Software Delivery Performance Benchmarks, 2023 State of DevOps Report

Where To Start? 

So where do you even begin? First, take a deep breath. You can’t change how an entire organization runs overnight. It’s about relentlessly marching forward with consistent, small steps. 

So the first small step? Check out DORA DevOps Quick Check to get a feel for where you’re at today. The 2023 State of DevOps Report and Accelerate are also worth a read too if you have a bit of time.

From there, just take one additional small step forward by trying something. Maybe you have one team that’s especially interested in these metrics. Great. Challenge them to think about how they could start assessing some of these metrics on a quarterly basis. Choose one metric that needs some work and experiment with ways to improve. What did you learn? And more importantly, what’s next? 

Keep marching forward. Keep learning. Repeat. 

And, if you hit any roadblocks along the way and want to brainstorm, reach out. We’re always happy to help.