Quality in Eurofins Genomics is a central focus point - analysis we do or products we produce have critical applications, be it production of drugs, identifying rare diseases or gene editing. IT is a driving force behind the scenes which challenges us to ensure the highest quality standards without compromising on speed.
When we start a new project, we do it with enthusiasm and feeling of doing something meaningful or even cool. Following scrum we quickly establish our velocity and deliver soon first release into production. Overall quality is quite good; results from testing acceptable, deadlines are coming so nothing can stop us. Let’s prioritize last bugs, fix critical, move rest into backlog – now we can be proud of having delivered value to users!
We continue delivering at ever increasing speed as team matures! Unfortunately the idyllic scenery gets soon destroyed by first, more and more effort needs to be spent addressing issues from both QA and production. We spend time arguing with QA and users on what is bug or if this defect is P2 or P3 or can even be seen as P4, from time to time we take a sprint to “stabilize”, but all too often nothing changes. User stories are getting spilled to next sprints, we postpone releases to have more time for testing, club them with next releases and finally find ourselves in downward spiral..
As quality cannot be compromised we quickly decide that Agile is fine, but as we work in regulated environment we need to be pragmatic and adjust Agile to our needs. What comes out is unfortunately not much different to Waterfall or V-Model, we still keep sprints and do reviews, but realize that only form is left. I am directly responsible for IT in Eurofins Genomics so will share experience from the field on how did we overcome this and reanimated Agile.