Software is never 100% defect free, so how far should responsible testers go before wading into the murky seas of cost overruns and diminishing returns—and what methods should they use to draw the line?
| Whether you’re reviewing a book for typos or testing the last batch of code before a launch, getting 100% defect-free is far too expensive to be realistic. Test too much and you overspend on diminishing returns; test too little and you risk embarrassing your company, angering your customers, and having to re-release the whole thing all over again. |
Am I the only person that seems to notice every mistake that makes it into print?
Whether they’re nestled within the pages of a magazine, newspaper, or book, my scanning eyes always seem to find every error. No matter how small or insignificant they are, they’re never enough to escape my gaze.
It used to drive me crazy. Every time I’d spot a mistake, I’d ask myself: “How can people paid to find typos miss such blatant errors?”
But that was before I started working in the software testing and QA industry. Now I understand that those overlooked mistakes aren’t borne of incompetence, but necessity.
The truth is software testers and editors have a lot in common: we take other people’s work, run through it with a fine-tooth comb, and identify any mistakes we find. We’re kindred spirits, ensuring the quality of someone else’s work—and by extension, the company it represents.
And whether you’re reviewing a book for typos or testing the last batch of code before a launch, getting 100% defect-free is far too expensive to be realistic. Test too much and you’re overspending on diminishing returns; test too little and you risk embarrassing your company, angering your customers, and having to re-release the whole thing all over again (with added schedule and expense).
Unfortunately, knowing how much is enough takes more than just going with your gut. And considering that 96 percent of major software projects will either go over budget, off schedule, or get cancelled altogether, I’d say a lot of IT departments are having a tough time striking the right balance.
When we first got into this business, most of our clients were struggling with that too. That’s why we developed a metrics-based, predictive approach to test planning that shows us how many defects we can expect to find in a given system or sprint. That gives our testing teams an established baseline that they can measure their defect-discovery progress against where they should be. The upshot here is unparalleled visibility—which translates to a clearly-defined endpoint that optimizes quality without overspending. It’s a core component of our True North Testing Methodology—and a big reason that we’re able to consistently deliver high quality results on time and on budget.
If your testing/QA team is having a hard time hitting project goals, we’re always happy to help. Whether you need help on a project/long-term basis, consulting work, or just a little free advice, don’t hesitate to give us a call or drop us a line.
After all, isn’t it about time your testers got started with the end in mind?
Cheers,
Mike Hodge
Lighthouse Technologies, Inc
Software Testing | Quality Assurance Consulting | Oracle EBS Consulting
