For example my stats for Feb will always be wack when compared to to July because of the extra 2 or 3 days.
How about weekends? My website's customers are office based so weekends are virtually zero, so how do I qualify the stats for a month with 5 weekends (like May).
What about know holidays?
I know working on a four weekly basis will solve some of this but I'm asked for stats on a monthly basis.
Cheers,
Number of visits divided by the number of workdays?
Seems that would give you the statistic you're looking for.
Being in the Analytics industry, I can tell you from experience with many clients that it is pretty much the de facto standard to allow a certain margin of error for this sort of thing.
The important thing to remember about web analytics is that it is trend based. There are too many variables and uncertainties and margins of error to try and nickel and dime things like this. You will get the most value out of web analytics by looking at trends over time, that is the rule of thumb to live by in this business.