The whole mindset of Continuous Delivery aims to delivery software with consistent quality. Automated processes encourage the usage of quality verifications.

Good software quality, of course, does not come for free. Sufficient test cases as well as code quality analyses require a certain time and effort. Automation and continuously improving the quality, however, will, after an initial threshold, pay off in the long run, and eventually lead to better software.

New features as well as found bugs need to be verified sufficiently during development in order to ensure that functionality works as expected. By automating the tests and keeping them as regression, developers can be sure that no new bugs can disrupt the functionality in the future. The same is true for code quality analyses. Once the analysis is set up with appropriate rules and the found errors are eradicated, it ensures that no new violations can find their way into the software. If new false positive violations emerge, the rules are adjusted and will prevent new false positives in the future.

Introducing new test scenarios, such as end-to-end tests, also highly supports this approach. Regression tests decrease the risk of newly introduced bugs more and more. Again, automation is the key. As we will see in Chapter 7, Testing, human intervention is helpful for defining reasonable test scenarios. However, it is crucial to the software quality that these test are then automated made part of the pipeline. By doing so, the quality is improved more and more over time.

This, of course, requires the engineers to put a certain priority into quality improvements. Improving software quality, as well as refactoring, doesn't provide any immediate benefits for the business. These efforts will, instead, pay off in the long run - by still being able to produce new features with a constant velocity or changing existing behavior with certainty that nothing else breaks.