The Automation Litmus Test

I’ve had this same conversation a couple times in the last week so that is usually a sign I need to write it down. The conversation usually starts with something like

This is hard, but would be easy if there were ids on elements or if there was a proper synchronization latch

Will the development team and/or management help the automation effort by putting in the hooks we need? Or will they shrug it off as not being important and not value producing? Of course, we know the latter means script development will be longer and that there are likely going to be more failures due to ‘things changed in the html’.

This is a critical point in an automation project, and one as a consultant I have labeled the automation litmus test. Automation is a full team project. Its not just the role of the scripters. If the company is not going to give the scripts the tools they need to be successful, the automation efforts should stop.


Have the scripters revert to being exploratory testers (you do have testers doing your automation, right?) and find problems that way.

When I am at companies doing something resembling traditional training I arrange it so that there is a tester and a developer working together as a pair. Not just all the testers on the left and the developers on the right. This way the developers can see first hand why having ids leads to tight locators and why latches immediately solve ajax synchronization.

I have the advantage of being an outside ‘export’ to bring this up with management; you might not be in such a position. But email me if you need me to talk to you boss. I’ll be nice. I promise. :)

Trackbacks & Pingbacks 1

  1. From A Smattering of Selenium #107 « Official Selenium Blog on 01 Aug 2012 at 6:25 am

    […] The Automation Litmus Test is one that most companies face, but now it has a name […]

Post a Comment

Your email is never published nor shared. Required fields are marked *