April 2006

Strategic Direction for Functional Test Automation

After many years of buying and trying to use test automation tools for functional testing, we must admit that our industry has totally failed to realise the predicted benefits. In the words of Paul Herzlich, Ovum Software Testing Analyst 'Functional Test Automation is Broken'. Today, only around 20% of functional testing is automated.

Is our dependence on functional testers reducing? The trend seems to be that developers will increasingly take on more unit testing and business analysts are starting to use a new breed of functional test tools – Test Automation Frameworks.

If we can't lead automation, where is Functional Test Automation leading us?

Facilitated by Susan Windsor, Managing Director WMHL Consulting

Test Process Improvement is a Waste of Time

Everyone wants to get better at testing, don't they? Everyone in IT wants to get better at everything, don't they? The usual approach from specialist consultants is to look at process. If we improve our process, we must improve our performance, right? Better, faster, cheaper - in whatever discipline we choose to focus on. Tools can help, training can help, infrastructure can help, but process comes first. But why is this?

The process improvement bandwagon started with Deming and others in the 1940s and 50s. The mantra then was, 'improve your product quality and cost of production by improving your processes based on statistical data'. This mantra works, but it works for manufacturing processes, with very high usage, mainly automated. Few people use production line processes to create software, so why do we delude ourselves that process is so important?

Perhaps it's because process is easy. Processes can be defined, scoped, taught, automated, measured, replaced. But what are the real barriers to improvement? People, culture and organisation. This session will discuss the potential for testing improvement approaches that might work.

Facilitated by Paul Gerrard, Gerrard Consulting