EverQuote’s engineering leadership team needed to see and quantify how teams were performing in relation to product development. They had previously attempted to gain these insights from Jira, but they couldn’t clearly connect that data to engineering work. They had tried another software intelligence platform, but they found the metrics weren’t as comprehensive or holistic as they needed.
A lot of important questions remained unanswered when it came to understanding how much technical debt was accruing and how fast PRs were moving. They clearly had bottlenecks in their code review process, which they heard in every single sprint retrospective, but they had no visibility into where or how work was getting stuck.
Velocity allowed EverQuote’s engineering leaders and managers to see the precise length of time that a PR was open, and how much time transpired before it was reviewed and eventually merged. They could also directly link each PR back to the relevant Jira ticket in
the same view. They discovered that when they were only looking at how many days an individual was coding, they weren’t accounting for everything that went into completing a Jira Issue. With metrics corresponding to each phase of development, they could gain a more complete understanding of Cycle Time, which provided deeper insights.
By taking a high level look at Code Review, they observed that several teams were approving over 50% of their PRs without any comments. With that data, they asked what type of work the teams were performing that didn’t require comments. It turned out that these teams were handling a lot of mundane tasks that were primarily configuration and did not really need review. That led EverQuote’s engineering leaders to advocate for more automation in their processes. As a result, they worked with product managers to automate more basic configuration tasks and reserve engineers’ time for more complex work.
EverQuote’s engineering leaders now have concrete data that they can bring to their CTO to help better diagnose the root cause of issues, inform process changes, make recommendations to teams, and ultimately, to evaluate the effectiveness and impact of those changes.
I love how I can mix and match a bunch of metrics together to look for correlations, make hypotheses about what’s happening with teams, and identify datasets that line up with those hypotheses to validate them.—Pheak Yi, Director of Engineering
Increase in PR Throughput per Contributor
Decrease in Cycle Time
Decrease in Time to Merge
Get more engineering leadership content, delivered to your inbox.
Get more articles just like these delivered straight to your inbox
Stay up to date on the latest insights for data-driven engineering leaders.