What is the minimum number of testers for programmers?

Question

I want to know if there is an ideal number of testers for each programmer. I heard that in many enterprises there is 1 tester for 20 programmers, and I can not imagine how this is possible, no matter what question exists, the ideal number of testers for each programmer? if exists, how is the relation?

Answer 1

It completely depends on the organization and what priorities they set during testing, how many testing units the developers themselves test, etc.

Answer 2

Depends on the programmer. If you work at Microsoft, the ratio should really be 20 testers per programmer.

Answer 3

In a ball from one to one to five to one, tester developers are what I'm used to seeing.

It varies by industry, team, development cycle, and many other variables. I would become heavier on testers with a less experienced or offshore developer.

This helps a lot to set up automatic tests on smoke test new builds or move on, as well as perform automatic testing and continuous integration.

Answer 4

it depends on the organization of the organization. But for a quality product, it should be 1: 2, i.e. 1 tester for 2 developers.

Answer 5

If you are doing test development , then there are 20 testers for 20 programmers.

Answer 6

It depends a little on what methodology is used, as if someone followed the pure methodology of the waterfall, then testers may not be needed, for example, until the next stage. Some enterprises do not have testers, although I think this may reflect other methods of quality control.

Where I work for the project, we have 5 developers and 2 testers, but earlier in the project we did not have testers. The quantity can vary, and I doubt that there is an ideal that can be applied universally to all sizes and methodologies of the company. Where I worked to where I am now, in most cases there were several developers and non-testers, and one tester appeared as often as the company hired a contractor, who then could use other opportunities in a few months.

In response to a question with a headline, I think that zero testers will be minimal, since it is difficult to imagine the presence of a negative or imaginary number of testers that would be below zero.

Answer 7

Of course, the minimum number is 0 testers. I think this is too little for most organizations. Most of the places I worked in were in the range of one test for ten developers, up to about one tester for 5 developers. However, we could always use more good testers.

Answer 8

This is possible because programmers are forced to test their own code before passing it to the tester ... but this is not always the case;)

Answer 9

You will not find a specific answer due to the huge number of variables here.

Outdated code can be very stable or very erroneous if it is enabled or updated.

New code may be erroneous due to programmers or even when it worked on code. I expect code made during the holiday or Friday to be more likely than problems during a normal work week on Wednesday.

It’s best to pay attention to your organization’s needs regarding available resources and make an informed decision about testing attempts. While QA is often given a low priority, it should not be, and a dedicated team (even if it is a team of 1) should check how everything interacts at all stages of development. Having done this, I can confidently say that it ends up costing less if a problem is detected at an early stage, and many warnings are given to programmers.

Answer 10

Like programmers, star testers can be much more productive than the bulk of their peers (the best guy I worked with was able to make almost half of the total number of problem reports on my own in some projects), if you don’t have such a star In order to have less than half the tester for each programmer throughout the project (from design to release), the risk will reduce the quality angles.

Answer 11

Programmer testers are not going to work. Try function points for each tester or functional subsystems for each tester. Testers are verifiers of functionality, not sieves for crappy code.