Please can someone help me understand something.
i have this chart when i have group by browser the number is higher than when i don’t have the group by on. Is there something i am not doing right? I can’t understand why the overall number would change.
If i take of the segmentation then the number is the same with or without the group by. I don’t understand why the filter is making it behave differently.
Best answer by jonathan
View originalHi
How much of a difference are you seeing when you group by browser? The browser property specifies not just the browser type, but also the version. Your report filters on the last 30 days. If a user were to visit your site, upgrade their browser, then visit your site again, that user would appear twice when grouping by browser. You can group by identity and browser to see how many users appear more than once.
+3
This is a flavor of a problem known as The Hotel Problem.
Uniqueness is a function of user identifier (User ID / Identity), dimension (Browser / Device Type / etc), and time period (Day / Month / Entire range / etc).
In your analysis, if a user visits from multiple browsers, or different browser versions, in the period you’re looking at, they would be counted multiple times.
You can’t sum users across distinct unique buckets or you inflate numbers as you’ve observed.
Hope that clarifies things!
Thanks for clearing this up it’s super helpful, we used to use GA and this sort of data was invisible to us without dev time, so now i can actually do it, the questions appear. Thanks agian.
Mike
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
