Whether you’re building your initial business case, calculating your ROI, or kicking off your first project, estimating the complexity of each test case you’re going to automate can be a real pain—which can make it near impossible to accurately scope your effort. Fortunately, all you need to do is remember 60-30-10.
|
Poorly-scoped effort initially might not seem like that big of a deal, but the unrealistic expectations it can cause in upper management can be potentially catastrophic. |
Let’s face it: test automation can be pretty intimidating.
From licenses to vendors, people to processes, automation is both a powerful tool and a significant investment. But while it offers a host of benefits that make it more than worth the effort, getting started can be tough. Sure, a lofty ROI is exciting and all, but it doesn’t change the fact that there’s a lot to do—and even more to prepare for.
And when you’re working in an unfamiliar field, it’s the simplest things that can cause the nastiest headaches.
Take scoping your effort, for example. Seems like it’s a pretty straightforward operation, doesn’t it? Just find some best practices for an estimate of how long it should take to automate an individual test case—and then multiply it by the amount of test cases you’re planning to automate. Sounds like a reasonable assumption, right?
There’s just one problem with that approach, though. It assumes that all test cases are equal—which any software tester will tell you is decidedly not the case. Now, poorly-scoped effort initially might not seem like that big of a deal, but the unrealistic expectations it can cause in upper management can be potentially catastrophic. What happens when your pilot project goes off without a hitch, but your boss gets cold feet because the numbers aren’t as rosy as the initial forecast?
Fortunately, there’s no substitute for firsthand experience—something we’ve got in spades. In fact, it’s how we discovered the secret to a fast, reliable automation plan is as simple as “60-30-10”.
Here’s the skinny: for every 100 test cases you identify for automation, you can figure that around 60 are likely to be easily automated, about 30 will probably take a few hours more, and 10 or so will likely take a few hours more beyond that. Naturally, the numbers will vary a bit, but this is a solid measure that’s held true throughout all of our engagements.
Over the years, we’ve accumulated our fair share of these little tips and tricks—and while we aim to share them all with you, sometimes things get lost in the shuffle. So if you’ve got a burning question or just need some simple advice, please don’t hesitate to drop us a line. Good, reliable predictors like these are essential for project budgeting, and understanding exactly the kind of investment you’re getting into—the last thing we’d ever want is for you make those kinds of decisions alone!
After all, isn’t the world intimidating enough already?
Cheers,
Mike Hodge
Lighthouse Technologies, Inc.
Software Testing | Quality Assurance Consulting | Oracle EBS Consulting
