I was taking a look at our traffic stats between the two services and there is quite a large discrepancy. Does any one have an idea about why that might be?
Probably because they are measuring different things in different ways. There is a saying that there is lies, damned lies, and statistics...
Site impressions are easy to measure so they should be about the same, but things like unique visitors and visit time is a lot harder. There isn't any way of identifying a user that works a 100%, and you can only measure when someone loads a page, not when they leave it, so visit times for exit pages are approximated.
"There are three kinds of lies: lies, damned lies and statistics." - Mark Twain's Own Autobiography.
Have you disallowed your own site visits by IP or cookie?