As we know, statistical data is to measure the success of programs and to identify areas of continuous improvement. But not every organization can be successful in addressing these objectives. Over the last few years, the number of client requests pertaining to establishing meaningful metrics programs has quadrupled, and it is bound to increase even further. While looking at several metric reports put together by client organizations, we found that:

1. Most testing metric reports were internally focused.

2. Every metric report had several months of data, with minimal analysis and no actionable items for stake holders.

3. 99% of the action items identified were related to the testing organization.

One of the primary reasons for these programs to fail is that most testing metrics are internally focused. To ensure success of testing metric programs, organizations have to clearly define both the type of metrics and also need to have a clear roadmap for the program implementation, data gathering, data analysis and reporting mechanisms.

Source: http://searchsoftwarequality.techtarget.com/tip/Software-testing-Why-test-metrics-programs-fail