The irony of TMM Henrik Emilsson

At one time, I worked as a tester at a company that claims to be at TMM-level 4 (perhaps 5).
Two things struck me:
Firstly, the quality of the actual testing performed on this company was not as good as one could expect from a company that have produced software for over 20 years.
Secondly, at the project post mortem, the testers complained about the lack of respect from the rest of the organisation.

One of the issues with TMM is that it says nothing about the quality of the testing.
A maturity level in TMM is looked upon as “a degree of organisational test process quality”. But it does not say anything about the degree of test process quality; or perhaps more important, test quality.
I think this has to do with the fact that many who promote and implement TMM has little knowledge in actual testing. This means that if you are to improve the actual testing, or tester’s ability to adhere to different test processes, you need to have knowledge about testing and people.
Or put in another words: it is easier to promote a single model than to improve testers’ skill.

Another issue with TMM is that when it is implemented, you might think that it is some sort of assurance that proper work is done constantly. This is not correct.
Since it is not saying anything about the quality of the actual work performed, your colleagues will not judge you on the basis of how high you have reached the TMM stair. Since “quality is value to someone”, you will be judged by your actual work – your performance. Therefore you cannot expect to be respected only because your company has reached level 4 on the TMM.
That is, respect is something you have to earn.

——

The maturity levels, according to the TMMi foundation, can be found at:

http://www.tmmifoundation.org/downloads/resources/TestMaturityModel.TMMi.pdf

5 Comments
Martin Jansson February 11th, 2008

Nice article Henrik!

One thing that I’ve noticed when talking about CMM and TMM. Some bigger companies use these models to classify their potential partners. In a sense it is very bad, but looking at it by strictly focusing on maturity of a subset what is presented in the CMM/TMM of the company/organisation to some extent it might shed some light to how mature a company actually is. Naturally this does not give the full picture and as you describe it can give a very wrong one saying that you are TMM level 4.

In either case it is easy to stay that these levels does not give the full picture. It is uncertain if you can assume much anything based on a statement of a certain level.

Henrik Emilsson February 12th, 2008

Thank you!
I get your point, and I think that is worrying.
So how should one point out the flaws in this method that they use to classify their potential partners? I do not know…
But if it is possible to question the method, e.g., by asking the “big company” questions about how they can be sure that what they look at really matters, perhaps new ideas on how to assess a potential partner might be discovered.
That is, if you ask questions on what is important and that give fruitful answers, it will soon be obvious to both you and the “big company” on what really matters.
The worrying thing is that if all parties take these methods for granted (like saying “a bad measure is better than nothing”) and not question the relevance, it will soon become a truth that isn’t true.

Rikard Edgren May 12th, 2008

I think the problem is that we interpret the numbers in the intuitive, but wrong way.
It is easy to think that higher maturity should mean better development/better testing.
All it really says is that this company has documented their processes according to a standard.

Henrik Emilsson May 12th, 2008

I believe you are right, Rikard.
Still, one major issue is that people are lead to think that this really says something about the quality of testing/development.
I think this happens because it is one major sales argument.
Which means that those who pay for it expect this to happen.
And those who are affected are told that this is to be expected.

Hi,
Thanks for the post.

Please update the latest link:

http://www.tmmi.org/pdf/TMMi.Framework.pdf