Hi community, I have a segment (at the hit level) that contains 3 URLs. When I breakdown the segment with the URL dimension the visits to each individual URL is higher than the total (EX: below it's 30+9+5 = 40). When I remove the URL with 5 visits, the total drops 36 (now reading like 30+9=36). I understand there are nuances with visit-level metrics that can sometimes get muddled when used against segments, but I can't seem to find an answer as to why this happens? What's going on and how can I articulate it to other stakeholders?
Solved! Go to Solution.
Views
Replies
Total Likes
There's nothing wrong with this. Some of those URLs are part of the same Visit, the total is the proper de-duplicated Visits.
So let's look at an example:
Visit 1
Visit 2
Visit 3
Visit 4
Page Views | Visits | ||
Pages | 11 | 4 | |
Page A | 4 | 3 | |
Page B | 2 | 2 | |
Page D | 2 | 2 | |
Page C | 1 | 1 | |
Page E | 1 | 1 | |
Page F | 1 | 1 |
Now, let's say we apply a HIT level segment:
HIT
Page equals Page A
OR
Page equals Page B
OR
Page equals Page C
Page Views | Visits | |||
Segment A, B OR C | 7 | 3 | ||
Pages | 7 | 3 | ||
Page A | 4 | 3 | ||
Page B | 2 | 2 | ||
Page C | 1 | 1 |
You cannot add up all the individual rows of data in Visits, or you will significantly over-count your actual Visits... in the non-segmented version you would have seen 10 visits (but there are clearly only 4 actual visits), and the same on the segmented version, you would have seen 6 visits (instead of the actual 3 visits that were returned by your segment).
In my example, 3 Visits match your segment criteria, Page A was seen in all 3 visits, Page B was seen in 2 of those visits, and Page C was only seen in 1 of those visits.
There's nothing wrong with this. Some of those URLs are part of the same Visit, the total is the proper de-duplicated Visits.
So let's look at an example:
Visit 1
Visit 2
Visit 3
Visit 4
Page Views | Visits | ||
Pages | 11 | 4 | |
Page A | 4 | 3 | |
Page B | 2 | 2 | |
Page D | 2 | 2 | |
Page C | 1 | 1 | |
Page E | 1 | 1 | |
Page F | 1 | 1 |
Now, let's say we apply a HIT level segment:
HIT
Page equals Page A
OR
Page equals Page B
OR
Page equals Page C
Page Views | Visits | |||
Segment A, B OR C | 7 | 3 | ||
Pages | 7 | 3 | ||
Page A | 4 | 3 | ||
Page B | 2 | 2 | ||
Page C | 1 | 1 |
You cannot add up all the individual rows of data in Visits, or you will significantly over-count your actual Visits... in the non-segmented version you would have seen 10 visits (but there are clearly only 4 actual visits), and the same on the segmented version, you would have seen 6 visits (instead of the actual 3 visits that were returned by your segment).
In my example, 3 Visits match your segment criteria, Page A was seen in all 3 visits, Page B was seen in 2 of those visits, and Page C was only seen in 1 of those visits.
Thank you explaining in very detail @Jennifer_Dungan to add to that, In simple terms, out of those 40 visits to the 'cybersecurity - consumer confidence' segment, multiple URLs are accessed in the same visit. So if you are looking at the segment level you will see 40 visits, but in the breakdown by URLs you will see the visits specific to the URLs, which won't add up to the aggregate visits of 40 as there are multiple URLs accessed in a single visit. Hope this helps.