I have set up custom variables a couple of months ago. Basically it tracks categories of pages, much like the example in the docs.

Now I want to get a list of the most popular categories. I've checked the dashboard, and the data is being reported fine. However, I noticed that on the dashboard, the amounts listed in the column 'visits' don't match the amounts I get when requesting the Analytics API. Instead, these numbers seem to correspond with the number of unique events.

Maybe the dashboard uses the wrong parameter here, it clearly says the visits column represents:

The number of visits to your site with one of more of these custom variables specified.

Or maybe I'm just interpreting this the wrong way. The pages that track these custom variables call two actions: _trackPageview and _trackPageLoadTime. All other pages call these actions as well, without the custom variables.

Below is sample output from the Data Feed Query Explorer for my categories: enter image description here

I want to know what's best to use in my API calls; the ga:visits metric is substantially lower than the number of unique events. How do I need to interpret this table exactly?

1 Answer 1


When looking at the definition of "unique events", it states that it is "The total number of unique events for the profile, across all categories." It is entirely possible that one visitor is triggering more than one event, especially if they are visiting multiple categories. Also, when you compare "visitors" with "visits", you should keep in mind that "visits" will be recorded only on those pages where the custom variables are being used.

I'm not sure if your example that you listed above is showing a different category in each row, or if it is the results for one category across multiple days. If each row is a different day, then it would be interesting to see what your numbers are by category to see if the unique events count for each category is the same. If each row is a different category, then it would be interesting to see which "unique events" are being recorded for each category.

Either way, the bottom line is for you to determine which piece of information is more relevant for you. Since it appears that you want to track information for each category separately, I would believe that using "unique events" would not be the best approach because those numbers seem to represent the events across ALL categories. You could also look at each number separately for different purposes.

For example, tracking the number of visits for each category will give you an idea as to the amount of traffic going to that category. Tracking the number of events for each category will give you an idea as to how much activity is going on in each category once you get a visitor there. Both approaches are relevants and provide important information. If the number of visits is low, it may be because the category isn't interesting or isn't promoted as well. If the visits are high but the unique events is low, then it could indicate that there isn't much going on in that category once people get there.

Don't limit yourself to one or the other. Using both in conjunction will give you a more complete picture.

EDIT - I found a link that provides a little more information about this. Since we can't see how you have your tracking code set up, it's hard to determine what you might be missing. Please note the paragraph pertaining to "unique events" on this page.

  • Thank you for your detailed answer. The rows in my screenshot represent categories. The thing is, sometimes the number of visits is 0, while the number of unique events is > 0. The other thing I don't understand is why the number of pageviews and unique events are different in my case, since I'm only tracking page views, that should be 1 event, right? Shouldn't these metrics be equal here? Aug 25, 2011 at 7:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.