Measuring coverage does not increase quality

We often use coverage as if it were a quality indicator… but do we really know what we're measuring? Code coverage, requirements coverage, branch coverage, decision coverage?

The problem is that when we don't have this clear, we end up making decisions based on empty metrics. Tests that increase the percentage but don't detect errors or provide value.

Coverage shouldn't be a goal in itself, but rather a tool to reflect on what we haven't tested yet.

In the video I explain why chasing a number can make us lose focus, and what to do instead to make our metrics truly useful.Retry

Frequently asked questions

Is coverage a quality indicator?

Not on its own. If we aren't clear about what we measure (code, requirements, branches or decisions), we end up making decisions based on empty metrics.

So what is measuring coverage for?

As a tool to reflect on what we haven't tested yet, not as a goal in itself.

Keep reading: What if the number your team celebrates the most is the one that's misleading you? and The most common mistakes when confusing testing with quality. If you want to go deeper, I recommend the course Risk management with examples from software testing..