How to deal with outliers when half of data are zeros?

Question

I have a dataset that shows the number of visits a user done during a year from each page.
For example:
0: means no visit from the page
27: means 27 times