I am creating an estimation for my testing and I have been wondering what are the universal risks that might affect the testing process. And what percentage should it take from the overall estimation?

To be more precise. What risks affect testing time directly. For example late builds delivery makes us stay idle instead of testing but does not actually affect testing time.

  • There are countless universal risks. My answer could list several hundred but I doubt it would help. The answer is that estimates and predictions tend to be poor - especially ones about the future. They also box you into meeting the estimate not the need for quality. – Michael Durrant Mar 15 at 16:39
  • 1
    Instead I would focus on the escalation process for when you find things. Will release dates be affected? How to escalate emergency fixes without finger pointing. How to rollback small changes. How to release more often, etc, etc. Traditional testing (verification after the creation process) tends to be a broken model. Modern testing is built into the development process and is part of its estimation (when estimation is done). Sometimes as simple as asking 'how and how long to test this?' during a sprint planning meetings. – Michael Durrant Mar 15 at 16:40
  • Actually, listing risks would help. especially if you really have a countless number of them. – Seeker001 Mar 16 at 9:44
  • there are thousands. I don't have the time and it wouldn't really add value as it would answer your question but not solve your problem. – Michael Durrant Mar 16 at 10:01
  • 1
    I'm not sure this question is answerable in its current form. If you think about it, literally everything is a risk, even things going correctly (after all, what are the unintended consequences of everything going "right"?) I'm not quite ready to close it by mod-hammer, but if I didn't have the mod-hammer I would vote to close. I think it needs to be refined, although admittedly I'm not sure exactly how right now. – corsiKa Mar 18 at 0:11

You need to know all the risks(imposible), but here are some of them. Like do you have framework, how much it is reliabile, how much experience on testing your team have. Is it risky change from development point of view, how much it affects application, how many TC you are expecting, does some TC alredy exist, are they automated, are they valid and realiable.False Fails are risky. Are developer able to deliver it to you on time (not last day of sprint). Does your testing require lot of set up (like for the dektop apps) if it is increse the risk. Does your app alreday have some critical bugs that prevents you to test. Do you need hard cooded DB values from Developers, and so on,..

  • This is a good answer. There is one problem though. Many of those risks like late delivery of builds does not increase the amount of time of actual testing. While builds are not delivered we are staying idle. I was thinking about risks that affect actual testing time. – Seeker001 Mar 18 at 9:07

Not the answer you're looking for? Browse other questions tagged or ask your own question.