Software tests have to specify the behaviour of your product as extensive and reliable as possible. At the same time their implementation and maintenance costs should be kept at a minimum.

As Kent Beck said before: “I get paid for code that works, not for tests”.

The Test Pyramid is a proven approach for this problem but leaves a lot of room for interpretation when elaborating a test strategy for your product.

Furthermore there are a lot of partly contradictory definitions of Unit, Service and Integration Tests and variations like “Test Diamond” and “Test Trophy” make it even more confusing. So, how do we get from the Test Pyramid to a concrete test strategy?

In my talk I will show you how to combine the principles of the Test Pyramid and the C4 Model for Software Architecture to obtain a test concept tailored for your product. Examples taken from a recent project will demonstrate how this approach balances test coverage, maintainability and development costs.