Or entry pages. We on-site searchers predominantly people who didn’t enter at the main page? Or the opposite?

Of course, you have to have an interpretation of what it means if they used on-site search. A lot of people think it’s a sign of confusion, so they would be a little worried if entering at an interior page meant people resorted to search. They’d maybe want to change the navigation available on those internal pages to be as good as what’s on the home page.

Or if on-site searchers tended to have a high proportion of a particular product type, they might wonder if that product type is hard to find using just the navigation.

Or you could use this filter for one of your KPI pages. Isolate people who signed up for something, and see if their entry pages were different, or their referrers. Or the keywords they used to get to the site or the keywords they used to search on once they’re there. Or did the people who signed up include a particularly cool site feature more than those who didn’t sign up? Maybe that’s an indicator of the persuasiveness of that cool feature.

I create a lot of these filters for KPI events or anti-KPI events (indicators of a negative experience). For each of these filters, I make one custom report that uses the filter as an Include, and another identical report that uses it as an Exclude. Then, using Excel, I compare.

The most common dimensions I use are Pages, Content Groups, Referrers, Search Phrases. Or custom dimensions based on a parameter that’s unique to the site, like ProductCategoryID or maybe Region.