I' ve been negligent on my blog due to a busy summer.  I spent 4 weeks in Beijing training our Protocol Testers and 2 weeks in Hyderabad training our Protocol Reviewers in Protocol document Quality Assurance Process (PQAP) which was an Honorable Mention at the Microsoft internal Engineering Excellence Awards. 

I’ve developed and submitted a paper for the Pacific Northwest Software Quality Conference (PNSQC) this year:

Too much automation or not enough?
When to automate testing.

Fundamentally test automation is about Return On Investment (ROI).  Do we get better quality for less money by automating or not automating?  The obvious and famous consultant answer is “it depends”.  This paper explores those factors that influence when to choose automation and when to shun it.

Three major factors you must consider:

1)     Rate of change of what you are testing.  The less stable, the more automation maintenance costs.

2)     Frequency of test execution.  How important is each test result and how expensive to get it?

I presented a rough version to Seattle Area Software Quality Assurance Group (SASQAG) last week.  

I highly recommend PNSQC as a great conference and a real value.  This year you can get an even better value by using promotional code of “FOA – Keith Stobie”.  This will give you a $50 discount off of the normal conference attendance fee.