I’m not sure what you mean by “improving testing”. That seems to cover everything else you listed and more –
good automation can improve testing,
project management is frequently (I hope) about improving testing.
Promotion to test architect requires scope of influence and just improving testing in ones own area is never enough. You must show you can help others improve their testing.
Project Management skills frequently come in useful to manage an automation project for roll out to a large group.
I agree with you keith that automation and project management are also aspects which improve quality. But I also think that you can have equal scope of impact by improving the techniques they use for testing separate from enabling more effective automation or better project management (they do all go hand-in-hand but I’m trying to distinguish between them). The root of the question was based on that discussing this with some other inidivual contributors at the PNSQC was that there seemed to be a general feeling that contributions in test methodlogy were valued less (in terms of career growth) when compared against equal impact contributions in automation or project management.
1) It is far harder to demonstrate that improvement came because of the techniques they use
2) It is hard to propagate the techniques (usually you add tools and projects J )
Determining “equal impact contributions” is just in general more difficult.
Let’s say you wanted to promulgate Model Based Testing or Pair-Testing bug bashes. You would have to show ROI, etc. versus current (perhaps undefined) methods.
MBT works best with tools tool support. Pair-Testing needs project management to succeed (I believe).
It is far easier to show how Tool X save Y time (implicitly give a good ROI). Project management is a skill managers are used to judging.