Traceability without the matrix

New Thinking 17 October 2014 Adam Howard

Traceability is an important notion – in order to give justification and credibility to the testing we do, we need to make explicit the relationship between the tests that we perform and the model that informs those tests. In other words, traceability is the relationship between a test itself – the what – and the reason we are performing that test – the why.

In traditional testing, the what and the why are typically documented as separate artefacts. The performance of a test is scripted into a test case, while the model of the software is documented as a requirement. Because those two things are distinct from one another, a third entity – the traceability matrix – is required to demonstrate the links that exist between them, to show that the requirements are covered by tests.

In Lean Testing, we develop Visual Test Models that combine the what and the why. This visually displays the software solution, drawing on any and all relevant information sources – from documented requirements to learned or sub-conscious domain knowledge – to create a rich model. Then, the actual tests that have and will be performed, and the information or learning that is generated, are plotted directly against the parts of the model they cover.

As such, there is no need for a traceability matrix in Lean Testing, because the relationship between what we're testing and why we're testing it is shown immediately within a Visual Test Model.