trending method for count data

#1
Does anyone have a suggestion on a charting/trending method for count data? I have reason to believe my data is not Poisson distributed because in most cases the mean is less than the variance. I would like to trend on a weekly and monthly basis. Is there any way to identify a week/month with unreasonably high counts?
 

JohnM

TS Contributor
#2
Sampled data from a Poisson may not be "exactly" like a perfect Poisson distribution, and may exhibit some sampling variation - I would still use it.