Question

To state the obvious, change management can be hard in software testing groups. From my experiences, this often appears to be true even when benefits to change appear obvious and non-controversial.

What successful case studies exist that can give hope to those of us who want to drive change initiatives and improve the testing of the groups we work in?

Answer 1

I just came across this excellent article written by Harry Robinson about a successful change management effort he was involved in at Microsoft. The paper is called "Obstacles and opportunities for model-based testing in an industrial software environment." Don't let the slightly drab-sounding title fool you; the content is thought provoking and impressive. Robinson uses a straightforward change management framework developed by Everett Rogers and clearly ties his specific experience at Microsoft to the following main points:

A Lens for Looking at Innovations Diffusion of Innovations is a landmark book by Everett Rogers that analyzes how new ideas spread through a group of people. His approach provides a useful framework for the issues we faced introducing model-based testing at Microsoft. Rogers lists five characteristics of innovations that can accelerate or impede adoption:

• Relative advantage: is your innovation better than the existing method?
• Compatibility: does your innovation integrate with the existing method?
• Complexity: is your innovation difficult to understand?
• Trialability: is it easy for people to experiment with your innovation?
• Observability: are the benefits of your innovation easily visible?

Robinson does an excellent job of using this "mom and apple pie" framework (that sounds somewhat vague, inoffensive, and posssibly obvious) to tie in specific tactics, strategies, and examples they used to increase adoption of model-based testing methods. Clearly, though, the piece has broader applicability to all of us who are trying to drive broader adoption of software testing approaches that we believe are under-utilized. I see a great deal of applicability in Robinson's article, for example, to the work I do with testing teams to teach them about using pairwise and combinatorial testing methods with tools like our Hexawise test case generator.

I highly recommend it as a concise, short read. Does anyone else have success stories to share?

Answer 2

I've never been one for case studies, but then I have never worked in an environment that they are needed. Usually I end up with the question "what is there to improve, and why?". If the same question put to a group elicits similar answers, then yes, you are onto something...if not, then it may just be you.

I look at change as something that people need to buy into, be a part of and want to accept. If you have resistance you need to play diplomat, not case study, because you need to convince someone who is already sure that things are fine that things need to change. A case study is not going to help you there, it may get you started and give you some ideas, but its not going to be the way to convince others. People Skills will help, and its been useful to me in getting things changed, the best part of it is asking a question and listening to the answer. That is step 1.