Best Practices

9 Metrics for Balancing Speed and Accuracy in Product Development

Better time-to-market is a priority for any development team, but it can come at the cost of accuracy and quality. There are many metrics to track progress and evaluate success in dev teams, but the following are especially crucial for those seeking to balance speed against accuracy and quality.

Cycle Time

Cycle time measures how much time is spent working on an issue – in other words, how long it takes the issue to “cycle” from one state to another (usually from open to closed). You can look at cycle time by type (how long it took a bug to go from open to closed) or by status (how long a bug sat open before it was closed).

Monitoring cycle time helps you set accurate expectations and identify bottlenecks or other pain points affecting your team. A spike in cycle time often points to a breakdown or a blockage somewhere in your process, meaning it’s time for an open conversation about what’s working and what’s not.

Release Cycle Time

This metric tracks your team’s total time per release: from the moment when a release began to when it was shipped. Tracking release cycle time can pinpoint holdups and process inefficiencies that delay delivery. Understanding release cycle time also helps you determine the accuracy of your ship dates and manage expectations around future releases.

Velocity

Velocity measures your team’s delivered value: the amount of value-added work, like features ready to test or ship, delivered during an agreed-upon time period.

Teams generally calculate velocity across three or more intervals to establish average velocity. Average velocity serves as a baseline for setting delivery expectations, understanding if your team is blocked and determining whether your process changes are working.

Your volatility is also relevant – that is, how much and how often iterations deviate from average velocity. High volatility signals that something is broken in your process or team.

Throughput

Throughput is your team’s total work output: the number of features, tasks, and bugs that are ready to ship or test within a given time period.

Usually, teams refer to last period’s throughput, but average throughput can be an especially helpful metric when it comes to capacity planning: comparing average throughput against the current workload tells you when your team might be overloaded.

Throughput helps measure your alignment with your goals. For example, if your primary goal is bug-squashing, you should expect to see plenty of defect tickets being resolved; if you’re focused on features, you should see more feature tickets than bugs. When your team’s progress is blocked, or their energies are pointed in the wrong direction, throughput drops.

Open Pull Requests

Pull requests are how developers ask their teams to review changes they’ve made to a code repository. This metric measures how many pull requests in a given repository —  or across all repositories — have been opened but not closed.

Your number of open pull requests will affect your throughput and iteration flow, so keeping an eye on this number can improve your team’s delivery time.

Tracking open pull requests over time also helps you spot bottlenecks in your feature delivery process.

Work in Progress

Work in progress (WIP) is the total number of tickets or issues that your team is currently working on. Like throughput, WIP is an objective measure of a team’s speed as a present moment indicator.

Ideally, WIP remains stable over time; an increase in WIP can identify inefficiencies in your process or signal that your team is overstretched (unless you’ve added more team members, of course!).

You can also divide your WIP number by the number of contributors on your team to get a sense of average WIP per developer. This number should be as close to a 1:1 ratio as possible, since too much multitasking tends to affect the quality and accuracy of your deliverables.

Iteration Flow

Iteration flow is a visualization of the status of your tickets over time. Flow is a good description, since this metric visualizes the shift of work from one status to another as your team moves toward completion through a defined process. It measures time to delivery along with the repeatability and health of your process.

Monitoring iteration flow helps highlight breakdowns in your process and get ahead of delivery problems, and it gives everyone on your team a tool to evaluate the predictability of their work.

Difficulty of Feature Implementation

This metric reflects your team’s perception of how difficult it was to implement a particular feature. An anonymous poll every sprint that asks your team to weigh in on the challenges of implementing the most recent feature(s) is a good way to collect the qualitative data you need.

Your team’s perception of the difficulty of feature implementation can have a huge impact on product roadmap and sprint planning. This metric helps you understand when your team would benefit from additional resources or extra training – or when they are ready to tackle more challenging work.

Should we release faster?

This metric represents your team’s perception of whether the team should speed up its cadence of releases. As with difficulty of feature implementation, you’ll need qualitative data to evaluate this metric. You’ll get the best information by polling your team every month or every quarter, since this metric doesn’t tend to fluctuate as much from sprint to sprint.

How your team feels about the pace of their sprints can give you insight into velocity, throughput and overall team health, while adding crucial context to your release window accuracy metric.

To get the full story on tracking data points for development teams — including more information on how to measure them and why ­— download our eBook, The Essential Guide to Software Development Team Metrics.