Many thanks for this article. I’ve been struggling with the transition to agile and the expectation that I am to automate tests, when I am not an automation engineer. I actually solved it by pairing up with an exceptional toolsmith who is taking basic Selenium recorded scripts (mainly recorded whilst I am doing my exploratory testing) and bundling the successful and ‘repeatable’ (for the expected regression) in with the development CI nightly build. The results are then put in to a DB and a dynamic web page is spun up so I can drill through all of the results. Metrics are produced off the top.

In essence this is what you are recommending and I thank you for putting some text around a theory I’ve been struggling to quantify. It works a treat and means I can spend much more time exploring the software and finding potential bugs.

Cheers

Rob

Me also following here this agile test automtion technique and we are highly
benefiting by getting the testers involved to know what they want and to
see things working in short time span,and basically keeping this pace ,parallely
we also have the massive automation developed,in small phases using agile test automation.After more than a year of automation,we had finally started to see success in keyword driven automation for the embedded automation which was once a night mare.I would like to join you in your research in this area if you do not mind.

- Kesavan

( You can reach to my works in testing, in google search )