Is software quality expendable?

This is one of tose questions that has been around for a very long time. A recent study from Original Software shows that it will probably stick around in the next years, as the world of CIOs appears splitted in two: 46% of CIOs think that quality has a very important role, while 41% sees it as a "nice to have quality" or even as a necessary cost.


I'm not a CIO (yet) but I have a very clear opinion on the matter.

Many CIOs have a special bend for asking "low cost working software, ready for yesterday". The problem is that the real cost associated with software is not writing it, but - even after a considerable amount of time - reading it for further modifications, bug fixing, evolutions, refactorings, etc. Leave that mess now and it'll cost you much more tomorrow. Software becomes kippleized (read this book for more infos on the term), someone says this happens even if you're not touching it and, mostly, when you're not watching. Bad quality software makes your technical debt increase, thus increasing the cost of software. Add the fact that most of the time you're modifying existing software rather than creating fresh new applications and you'll have an idea about the huge impact this debt can have.

And... is that low cost software really working? The forces involved in software development are intertwined, so you can't modify one without affecting the others. Literature normally refers to time, scope, cost and quality. Cut time and cost and guess what happens.

Another very important aspect, albeit disregarded by CIOs, is that developers love their job (well most of them do). We try to do our best, and we are proud of it. We are not satisfied when time is cut and cost and scope are kept fixed, as quality necessarily drops. Nobody wants to produce sloppy code. Nobody wants to be forced to produce sloppy code. Somebody might even think to move to another company, which, again, means a non neglectable cost. That could even start a trend, and you can see a company slowly transforming into a sinking boat. I'm probably going too far, but it is just for the sake of clarity.

The whole study can be downloaded here (some informations required).