How do I convince a customer to implement more tests than just "happy case" tests

Question

I have been building a test automation project for a customer that uses Microsoft Dynamics 365 as their business platform.

All test cases I have been building for them are happy cases, the chance for them to go wrong is very very remote.

• If tests do not catch any bugs, they serve no purpose but to provide confidence to the management that "happy cases" are working.

There are many many more tests we can potentially implement for our customer, such as:

• Negative tests
• Tests with mutation

But to our customers, hiring us to implement those tests will cost them quite a bit money, and that is probably the primary reason that stops them from doing it.

It is really hard to convince someone to implement a prevention when the problem has not happened yet.

Any suggestions?

Answer 1

Without knowing the details, if you as a tester see that the happy test cases will never fail, then the money invested in them is probably wasted.

Some thoughts:

• You might want to think of other ways to describe the tests that you are doing. For example do they cover all of the Boundary values and equivalence partitions. Happy testing is a popular way of making the testing process easy to understand, but is not a good way of justifying value for money.

• One way to present "Negative test cases" is to discuss them with the customer as "real world situations", which the application will experience with real usage, and could damage customer reputation.

• Put aside a little time for exploratory testing to hunt for bugs, this might uncover evidence of faults and encourage the customer to reapply their spending away from happy cases, to the areas the application needs improvement.

Answer 2

I recommend letting them know, in writing, that it is risky to skip negative testing.

After that, it may make sense to let them discover the outcome for themselves. If the outcome is painful, they may be more willing to invest in negative testing.