I recently had a discussion with a project manager who had just came back from a gate review meeting. The discussion was heated when it came to the testing aspects of the project. The project manager worked with the testing experts on the project team in this regard and had identified the number of parts that would be subjected and the variety of tests to be conducted. The committee did not like the numbers even though this number was 2x the size of a similar project done by another part of the company.


The project manager kept asking those around the table that were critiquing the test plan what they would prefer. To which those around the table told the project manager the PM should tell the committee members. The project manager's retort, I did already and it was from the experts that are on the project team. This kept going back and forth until finally, the committee came up with a 10x number of the original. This was level of testing expected by the committee, even though the experts counseled otherwise, and the launch window really precluded such a number of test specimens. There simply was too little time remaining to be able to have such a large set of test specimen. Additionally, the experts in this case had more than 2 times the number of test specimen that another part organization did for a similar project.

As I listened to the project manager describe the way this discussion unfolded, I could not help but think of a random number generator being used to determine how the product should be tested. The committee members did not know what "right" was, but knew what the experts were suggesting was errant. They wanted a new number from the project manager, and the project manager referred to the experts, the committee members seemed reluctant to answer but eventually did with an answer that would not meet the launch needs.