‘How to quality’

We use a 5-stage model created by Lisa Crispin, which considers quality from differing perspectives.

The first stage, Development-based quality, deals with the aspects that pretty much provide the basis for software development. For example, terms such as ‘test automation’, ‘testarting’, ‘pair programming’, ‘reviews’ or ‘static code analyses’ are covered here.

The next stage is Product-based quality. Does our software work as expected? Answering this question is supported by TDD, BDD, security aspects as well as load and performance tests, etc.

The third perspective is User-based quality. We take the customer's perspective here: what are our customer’s needs? Documentation and feedback sessions, UX and cross-device testing are examples of how we approach this.

Value-based quality is primarily about the cost-benefit factor: how much more can I do to achieve further improvement? Examples here are considerations of velocity, cost of infrastructure, simple onboarding for new colleagues, and also complexity.

In the last stage, Transcendent quality, we look at ‘transcendence’: this is hard to measure – it's about mentality and mindset, so to speak. Is the team spirit right, and do all team members share the same mindset? To give a couple of examples, we value retrospectives and treat bugs as top priority.

At the beginning of the Workshop we ask the participants for their personal scale-rating assessment of the quality of their team and its possible need for improvement. We repeat this query at the end of the workshop. This scale-rating survey is used to compare the development during the workshop or for a later re-run.

Following this brief assessment we explain the 5-stage model and move on to the working phase.

All participants are now given around ten minutes to gather everything they can think of in terms of quality and QA, and assign these items to a specific stage of the model. What does ‘quality’ mean and what does it include? Anything that comes to mind may be mentioned: methodologies, tools, suggestions, ideas and technologies, etc. – there is no ‘wrong answer’.

Following this we collect the whole team’s thoughts and assign these to the appropriate stage of the model. In addition facilitating the Workshop, here we are already trying to contribute our experience so as to include important aspects that may not have been mentioned yet.

The overall picture that has now been created represents the team's understanding of ‘quality’.

We cluster multiple or similar nominations and then ask about the strengths of the team. We can use these to create synergies and, if desirable, to network other teams with the know-how of the participating team.

Next we want to identify and work on areas of potential, which we do anonymously via a vote. Here we look at the most frequently mentioned topics first.

For these topics we now move into an open discussion, similar to a retrospective. We discuss the points, collect notes on discussions and try to work out specific measures. In this phase we contribute our own experiences and tips to the conversations, or we can also network and communicate, for example. Each measure should be assigned a deadline and an owner to make sure it’s driven forward.