Hi Jeremy (?),
I "kinda" have an answer, but not a very good one. I guess you will know what I mean when you see it.
First, I created a calculated metric that used the IF function and basically set the Visitors (but it works with any metric) to 999 billion when the original value is below x (150 in my example) and just copies the original value when it´s above 150.
Then I sorted the table by visitor (ascending) and limited it to n rows.
On the plus side, you actually get a table that shows the bottom n rows with more than x visitors.
BUT: when you sort the table with any other metric than "modified" Visitors, it goes belly up.
Actually, I don't like my solution very much.
If I were hard pressed, I probably would run a report in Excel using the Report Builder plugin.
I'm looking forward to more creative solutions than what I cobbled together.
I was under the impression that a segment is just a collection of visitor IDs, Visits or Hits that follow a specific definition. It is not a collection of other dimension items, e.g. products. (though that would be a great feature)
According to the Visit Number help page, it's the n'th visit of a specific visitor.
So, your segment would include all hits in visits that included a view of page "xyz" and are visits number 4, 5, 6, 7, 8, 9, 10 of that specific visitor. When I tried it in my segment builder, the dialog showed a kinda small number of hits/visits in that segment but that may be different when you have a site that is regularly used by the same visitors (twice a month or so, like a banking site).
Hi segments can be almost anything you desire...
Can be a page, with a a purchase event and for a specific product for example...
Like any filtering the more detailed and complex the segment is the smaller the results tend to be.
Also when building a segment Check under compatibility as some segment elements are not compatible with other parts of Adobe Analytics.
thank you so much.
I tried that same solution but set the <150 to 0 which means my sort logic didn't work. Setting the <150 to 999B would solve my problem, and I understand the limitation you referenced. for my purpose where I'm only looking to provide a 'snapshot' of the worst performing items, your idea will work.
Again, thanks so much!