If you have multiple locations, with different reporting units, determining what level of detail you want to look at the results of your mystery shopping program can be a challenge. Too many levels, and you can loose the effectiveness of your program. Why? It's easy to start blaming others, and not take personal responsibility when the numbers grow too large. Also, once the "big picture" gets zoomed out too much, you may not be able to view the forest through the trees. I'm not suggesting that you don't utilize a "roll up" report to measure work units performance. I'm saying you should have fewer details the more data that's included in a report. As an example, if you are the CEO of a company with 30 operating divisions, you may only want to see the "top level" average scores for each division, rolling into a company score. Each division manager needs their rolling average, plus their average scores for each section of the observation. Forget the fancy charts and graphs plotting each division against the other. They are not autonomous entities in the free world. They are you. If a customer goes into a location in New York and receives exceptional service, then visits another location across the country and is treated abominably, you lost. Plain and simple. Don't give managers "wiggle room" when discussing their numbers. If their scores are down, train a lot and shop more. If their scores are up, throw them a party and send congratulatory notes to every employee shopped. Remember, if your employees know you're looking at their results, they will pay attention too!