Track Talk, W11

If I Could Turn Back Time, Testing Differences I Would Make

Lloyd Roden

13:45 - 14:30 CEST , Wednesday 17th June

I have been in the test industry for over 35 years and have learnt many valuable lessons. Hindsight is a wonderful thing and if I had my time again I would do many things differently. Aspects such as:

* Test Estimation

As a Test Manager, I had always found estimating testing difficult and once I had created and presented the estimate to my management, found myself in endless arguments trying to defend the estimate. I would now do things very differently.

* Release Decisions

One of the worst moves of my career was to demand that I made the release decision. The release decision should never reside with the testers/test manager. We are a signpost – pointing the way. We present the facts and data regarding system quality to the stakeholders and they are ultimately responsible for making the release decision

* Test Plans

I firmly believe that the activity of test planning is vital on any project. Just like any major event in our lives; business trips, holidays, getting married, starting a family, retirement etc – we need to plan, otherwise problems could arise. The output of this activity in terms of a written document is less important, in my view, especially if it is not used.

* Test Reports

I have reported to many different levels of management and can honestly say that the higher the management the less information they want. As test managers we need to ask them what they want in terms of information and if they don’t know, then we start small and build upon that.

* Trend Analysis

I would definitely use trend analysis to try to predict what is going to happen. Burndown charts are a good example of this.

* More manual testing

I am a firm believer in test automation and am convinced that we need it more than ever in today’s IT environment. However, I have seen a trend that seems to be to do more automation and less manual testing. Test automation is not the best at finding the failures of the software. Manual exploratory testing is the most effective way of finding the failures.