I work for a web development agency, and we are using the scrum model.

The way it is applied may not be the best. We use grooming sessions, work in complexity points, but in the end, the project was still sold in dollars.

Unfortunately, in order to stay "on target" on a tight budget, we often end up cutting QA short (we do not have a department for that).

There is currently no automated testing, as it would take a lot of time to create them on the framework we are using.

So my question: how do you make QA, and the bug fixes that come with it, work well in scrum?


Scrum requires that the team creates a potentially shippable product at the end of every sprint. This means that all testing needed to release code to production for an item must be done before you can say the item is done. In other words, no partial credit for Dev-Done.

There are many reasons Scrum encourages, but suffice to say that if you are not creating potentially shippable increments each sprint, you are pulling out a critical part of Scrum and you will lose key pieces of value that it provides.

Now, many teams can't jump right to this and they shouldn't give up on Scrum. They should acknowledge that they are creating a deficiency - they are creating technical debt - that they will have to pay down sometime (bugs don't go away just because you aren't testing for them) and they should be continuously improving toward a way of working that allows for a potentially shippable product every sprint.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.