Why Do Top Software Testing Companies Prefer to Calculate “Rational” Test Efforts
Software projects that set a challenge to meet tight schedule require longer periods to be completed than they would do if they have begun with reasoned and realistic schedule. Tom DeMarco.
“There is not enough money in the budget to pay for testing! It is a pity but you will be asked to finish / stop before the target milestone.”
Actually, it is a practical and habitual way of cutting spending within each testing budget. Sure, it is not a very good example but rather harsh and little bit vague, however, it is quite common.
Top software testing companies have in-depth knowledge of software testing and huge experience in testing services but they do not ever attempt to meet aggressive project schedule though it is easy for them. And, as practice shows, any budget cuts lead stakeholders to nothing good, namely, this adversely affects the quality of the developed software. Moreover, it is hardly possible to complete any project within its customer’s short schedule. What is the reason for that?
This can be explained by the fact, that there are often unrealistic and biased approaches towards the effort estimations. Project managers guess that if they give aggressive (early, unrealistic) milestones it will help them to get promotion but it is not like that.
The other wrong practice is calculating test effort as the percent of time the development takes. Those who find this approach suitable for the intended purpose, first use it for their development effort estimation and only then resolve on an appropriate testing effort.
What do you know about QA outsourcing? Have you heard that it is possible to reduce development time and at the same time save on production costs? Outsourcing your qa and testing needs it will be much easier for you to achieve these goals.
Do not allow anyone to disorient you! It is false that testing effort is more closely related to anything else than to development time.
Nobody can be 100% sure that there must be 110 units of development effort for 95 units of testing effort or 100 or even 1,0000 units of testing effort if the estimates are not based on consistent foundations. Estimations of the kind are variable with respect to domain knowledge, code /application complexity, familiarization, testing types, software development strategy / method / methodology, team sizes and a lot of other factors.
Web, desktop and mobile app testing companies do not ever set to work before estimating real test efforts. As a rule, they choose to use the proven methodologies such as task-based estimation, Delphi analysis, test point analysis, requirement-based analysis, scenario-based analysis and so on.