Where to put the new Tester?
From the first tester in the team to the nth addition in the team, it is a very tricky proposition so as where you should be putting the new person. If you are on a layered product, you’ll be considering which layer has more priority: the lower most layer as it has the most functionality or the upper most layer as it will have the most user interaction. If you are managing more than one product, you might be considering whether you should put the tester on a newer/un-tested area or should put on a half-tested feature to complete the testing. If your product is getting hit by the non-functional requirements, you’ll be considering whether performance testing is more important or you should let the new tester attack the security layer. There is no thumb rule or easy answer to this question but here is one suggestion of a model that has worked for me: Do a Software Testing Heat analysis.
Well, I don’t mean the usual heat analysis here and don’t want to expose your product or your team to more heat. It is a way to look at how well your testing effort is scattered and it has some inspiration from Google’s ACC (Attributes Components Capability) model.
Depending upon on what basis you want to do the analysis, you can list down your sub-parts. If it’s a test department, the list can be products or projects. If it is a product testing team, the list can be features. For example you come up with 9 such things mentioned in my diagram below as A through I.
The next thing is to have a legend of colors so as how you want to represent the testing effort. For example ‘Green’ means well tested, ‘Yellow’ means some testing and ‘Red’ means no testing. ‘Orange’ means you are not sure. Now assign colors to your boxes and you get some thing like this:
Let’s take up the same question again: “Where to put the new Tester?”
The best places are the Red ones as there is no testing though you might be tempted to turn some of the Yellows to Green but my advice is not to do that. We were on a project many years ago, where we had a central component which was largely tested by the Programmers in the team and we had all the testers on the upper GUI layers. We got some bandwidth and spent on the ‘never tested’ area within weeks the Project Manager said to me: “I wish we had put a tester earlier here. He has been finding some really big issues in the platform”.
I have used the same sized box in the above example for simplification assuming all sub-parts are equal. Your boxes can be of different sizes letting you further select which ‘Red’ box is most deserving.
Do you have other suggestions on where to put the new tester?