Code quality is often seen as a badge of honor for a software system. Taken at face value, high-quality software should be a cause for celebration – and it usually is. After all, quality is a sign of a readable, reliable, and maintainable codebase.
But what if your data shows a ‘gold-plated’ codebase, but the delivery of new features is comparatively slow? It’s a paradox that fascinates BonCode CEO Jan Willem Klerkx, an experienced specialist in the fact-based measurement of software quality and delivery performance.
As Jan Willem explains in this blog, quality alone doesn’t tell the whole story. Churn – the measurement for how much code is actually changing over time – provides crucial context for quality. Without churn data, organizations risk mistaking polish for productivity.
“If your software quality is consistently outstanding, it’s worth checking the number of change requests actually reaching your dev teams.”
– Jan Willem Klerkx, CEO, BonCode
What is churn in software development?
Churn in software development refers to the number of lines of code that are added, edited, or deleted over time. It’s a basic but revealing metric that, when paired with quality data, paints a more realistic picture of what’s happening in your software teams.
Take this (real but anonymized) example. When an internationally recognized company commissioned a quality assessment for a major project, the results appeared pristine: top-tier architectural quality, what we would call ‘gold-plated’ code.
And yet, hardly anyone involved with the project was happy with it.
“Next to nothing was being delivered,” recalls Jan Willem. “The system looked amazing, but it wasn’t growing. When we analyzed historic versions of the codebase, we found that churn had flat lined. There was no movement. That’s when we knew something was wrong.”
In another real-world example from another global enterprise, a team of seven .NET squads delivered incredibly high-quality code – but very few lines of it. Despite being four months into the project, the growth rate of the codebase didn’t reflect the team size.
“The code quality was, by BonCode standards, ‘gold-plated’, but the system itself was small. Then the client realized something: the real problem was their requirement-setting process.”
– Jan Willem Klerkx, CEO, BonCode
Spot the signs of a stalled software project
We’ve seen the same pattern emerge across various industries and sectors, particularly in large organizations with complex governance. Here are three common red flags:
- Developers spend more time polishing code than producing new features: With no clear functional requests coming in, devs default to perfecting what’s already built.
- Change requests have dropped off: If dev teams aren’t being fed new work to be done, they can’t act, even when they’re technically capable and well-resourced.
- Management is misaligned: In larger organizations, there are often additional layers of meetings and managers, creating bottlenecks that extend or block delivery timelines.
Go from opinion to insight with BonCode
BonCode’s quality tooling and consultancy turn subjective concerns into objective insights. By combining code quality metrics with churn data, your organization can identify when projects are delivering real value and where things are stuck in neutral.
Jan Willem stresses that measurement isn’t about micromanagement. Instead, it’s about creating balance. “No one wants perfect software that doesn’t do what users need. And no one wants lightning-fast delivery that breaks in production. It’s all about balance between speed, quality, and output.”
Need to get your software projects back on track? BonCode helps you surface actionable insights, independent facts that cut through internal politics.
Find out how to get clear oversight of your software projects. Get in touch today.
