A more resilient review practice and a higher-throughput team

Redistribution tracked over time

Once Span quantified how concentrated code review was among the team, Mike recognized that the pattern as unsustainable and began deliberately redistributing review work across the broader org. Span's visibility made the rebalancing measurable: rather than relying on impressions of whether the load was actually shifting, they could track whether the same few engineers were quietly absorbing the work again, or whether the redistribution was holding.

This redistribution was viewed as a structural investment in the resilience of the org. More engineers reviewing more code often meant more knowledge sharing, more growth opportunities for mid-level engineers, and a review practice that didn't depend on any one person's availability.

As a result, weekly code review throughput rose from roughly 150 reviews per week in 2025 to 225 in 2026. This was a 50% increase achieved with a smaller team thanks to a more balanced distribution of work. 

Cycle time reduced 20–30% across teams

The same visibility that surfaced the review imbalance also surfaced bottlenecks elsewhere in the development workflow. Span made one long-suspected drag legible: Coalesce's CI/CD pipeline was costing the team meaningful engineering time. "We knew this in our hearts, because we felt it as developers, but now we knew how much time the team was being spent babysitting our CI pipelines,” explained Mike.

Coalesce dedicated two quarters of focused attention to cycle time, using Span to identify where work was getting stuck — whether in review cycles, CI runs, or merge queues — and to track improvement over time. The result was a 20–30% reduction in cycle time across the teams where the focus was applied. Combined with the throughput gains from the review redistribution, the team was moving meaningfully more work through the system at a faster pace.