Gary Flood talks to Colin Armitage, CEO of software quality and testing firm Original Software about the latest challenges in improving the quality of software in the enterprise
‘Application Quality Management’. Isn’t that just ‘testing’? I last looked at that on a regular basis in the 1990s. Has anything happened?
It’s true to say that while small players have come and gone, the dominant players are still probably the ones of ten years or more ago, though many have lost their independence: so you have Mercury [bought by HP in 2006], Rational which is part of IBM now [as of 2002] and recently  Micro Focus bought the Compuware testing portfolio which included the Segue application suite that had ended up in Borland.
That does seem to suggest a lack of dynamism in the market.
On the contrary, not only are we seeing growth, we also see a lot of market frustration. We have a lot of second-time buyers – people who have found what they are getting from the established vendors to not be delivering, to be too much hard work getting to deliver what they need.
Code, whether developed in-house, bought in, developed on- or off-shore, is inherently complex and very difficult to make fit for purpose…
[Interrupts]: Well, simple – I’ll get rid of the problem and get it off the cloud so that that complexity becomes someone else’s problem… Sorry if that destroys your business model.
[Pauses then laughs]: Back in the real world, that isn’t an option. The reality is that promising as cloud may be, so far it’s only the larger companies looking at it at all and they are not putting the applications that matter most anywhere near it either. That may come, but by stages. Meanwhile, the reality for most organisations is that they are a mix of everything, green-screen, mainframe and midrange, first generation client/server to Ajax.net.
So the claim is that your company offers something that IT shops need but they are not getting from the established testing suppliers or the cloud as-is. Aren’t the package suppliers playing their part? Isn’t their code up to scratch?
It’s more an issue around version control. An SAP releases a lot of product and it takes a lot of customisation by most environments to deliver what’s needed for the business, so there is huge, on-going effort to re-engineer, re-test and re-deploy that code. Yes – there is business for us. The total market for test management and functional and load/stress automation is currently valued by Gartner at just under $1.2bn, and is growing at more than 8% per year, for instance.
Tell us some proof points about Original, then.
In terms of our history, we were founded by in 1997, with an IBM midrange focus originally, but TestBench is now used on many, many platforms now. In terms of customers, I am extremely proud of our partnerships with organisations like Unilever, Euronet, Intuit, Pfizer Pharmaceutical (Ireland), HMV, DHL, Coca-Cola, Skandia, HSBC and many others – more than 400 organisations operating in over 30 countries use our software now.
Analysts Ovum have said that by sticking firmly to a value proposition almost solely around unsolved challenges in functional test automation we’ve both filled out areas other vendors have left alone but also made test automation much more accessible to non-technical testers. Gartner last year put us in the ‘visionary’ part of the Magic Quadrant it issues re the whole Integrated Software Quality Suite area.
Let’s sum up. What is your call to action to the CBR readership about application quality management?
Most CIOs see this issue as being the problem they set up the QA Department or function for. It’s not. Testing and the drive to quality has to become transfused into the lifeblood of the company, from QA of course but also development, the helpdesk, the users. There has to be an ongoing, conscious commitment to constantly improve and support the applications the business depends on. That’s because like it or not, this problem will never go away and so is all our job to work on.
What makes you so sure it won’t ever ‘go away’?
Anyone who’s worked in software for a while knows that there are always defects – all we can ever try and do is attempt to deal with the most severe and obvious. The challenge, indeed, is finding the most serious in the first place, often.