On his model-based testing blog, Michael Mlynarski writes:
In this session we have discussed the ROI of software testing techniques. More than 20 experienced testers or test managers had difficulties on formulating the ROI of any testing technique. Two main arguments were about reducing the costs for bug-fixes in late stages of the development process and about increasing the test coverage. […] On the other hand, increasing the test coverage was not a sufficient justification for investing in MBT.
(See the original post here.)
We have seen this same problem a multitude of times. The main business argument for switching to model-based testing is that it makes business sense. But if a company’s DNA does not really allow it to measure the difference between good, better and very good testing, it is impossible to put through the methodological and cultural changes that model-based testing requires (including using commercial test design tools and having to evolve engineers’ skills).
This situation will of course in the course of time (as with regards to so many other industrial innovations or even revolutions), but it’s difficult to predict how quickly. Our customers have teams who have used our automated test design solutions for years and have achieved a predictable, repeatable and consistent reduction in cost and improvement in quality, but still they find that trying to get the next team behind the corner to use similar solutions is next to impossible because of political and cultural inertia.