Symptom: People can't be sure that their quality is satisfactory day-by-day as they do their work.
Imagine an operator of a machine which has a dial and a knob. When the dial gets out of tolerance, the operator can adjust the machine.
Now imagine the dial moved to the supervisor's office, and the supervisor tells the operator how he did at quality at the end of each year. Of course, there's little chance of a quality product.
If people do not receive ongoing feedback through metrics of their performance, a formal program to implement metrics on a group-by-group basis might enhance performance.
Every group should have a set of in-process metrics that tell them how they're doing. The metrics must be comprehensive so that people don't optimize one objective at the expense of other equally important (but perhaps more difficult to measure) deliverables. While quantitative metrics are preferable, precise numbers are not as important as relevance, controllability, timeliness, and comprehensiveness.
These metrics must be delivered to the people doing the work, not their supervisors, to allow then to continually improve their effectiveness. And they must be delivered in time for the person to correct their behaviors.
