Segment showing UV counts much lower than by Page Dimension in Freeform workspace

[DeebyCF]

We rolled out a new version of a portal in our website back in June and I am trying to combine actions that are the same within that portal but were named something different before and after the switchover.  I have a freeform table that shows all the different page names before and after the switchover to the new version within the portal that we want to measure and looking at them by Unique Visitors for the last year.

I thought I might be able to create a segment that would look at each specific action in the portal we are measuring and it would combine the UVs between the two different page names from before and after the switchover for the same data range, 

However my UV numbers when viewing by Page Name in the Freeform Table are significantly higher than what each Segment is showing.

For Example

We have an action in the Portal named xyz:view:idcard before the switchover and after the switchover to the new version of the Portal it is named xyz:ss:view:idcard.

In the Freeform table for 2025 it has the Unique Visitors for 
xyz:view:idcard  = 200,000
xyz:ss:view:idcard = 100,000

So cool, 300,000 UVs have used the idcard action so far this year.

But if I put xyz:view:idcard and xyz:ss:view:idcard in a Segment that has
Page Matches xyz:view:idcard
or 
Page Matches xyz:ss:view:idcard
The total UVs displayed in the chart is like 85,000. I have tried setting the Include to Hit, Visits or Visitors and it the numbers are much lower.

Does Segments handle Unique Visitors differently?

[Jennifer_Dungan]

Since I can't see your segments, I am going to make an assumption based on your wording that your segments are using Visitor scope?

This is probably your issue...

I see you said you tried HIT scope... this should be the correct scope... paired with your UV metric in your table.

When you use VISITOR scope in a segment, it pulls back ALL hits from all Visitors that have ever match a page of "xyz:view:idcard"

I don't know if you original numbers 200K and 100K are from using Visitor based segments, but it's possible that you overcounted in your first numbers....

Let's look at a sample:

Segment:

VISITOR [

    page equals "x"

]

and then this user's visits:

Visit 1

• page x
• page y
• page z

Visit 2

• page a
• page b
• page c

Using the segment above, you will get 6 PVs, 2 Vs and 1 UV; since the page x in Visit 1 matches your criteria, and now all hits for that user will be included in the data that is being returned.

Using a Hit based segment:

HIT [

    page equals "x"

]

Will return 1 PV, 1 V and 1 UV.

While my article focuses more on include/exclude logic, it does touch a bit of segment scope:

https://experienceleague.adobe.com/en/perspectives/the-magic-behind-the-curtain-complex-segments

You can also watch the session I did for skill exchange last year that helps illustrate segment logic:

https://experienceleague.adobe.com/en/docs/events/the-skill-exchange-recordings/analytics/aug2024/se...

If you can share more info about the segments you used, we can help confirm if this is indeed a scope issue.

[DeebyCF]

Thank you so much for the info. I will look at the scoping

[Jennifer_Dungan]

You're very welcome.. It's easy to mix up the scope of the segment with the metric you want to look at (given the similar naming)

And @Vinay_Chauhan  is right too.. if you added two segments into your original workspace, the total wouldn't have been de-duplicated... It's strange that Adobe doesn't maintain the deduplication when using segments, but for both "real" segments, as well as quick or adhoc segments in the breakdown (i.e. when you you filter the results to a specific few) the behaviour shifts.

[Vinay_Chauhan]

Hi @DeebyCF 

What you are seeing is a common issue in Adobe Analytics and it comes down to how unique visitors are handled in freeform tables compared to segments.

In a freeform table, when you place Page Name as the dimension, the unique visitors are calculated separately for each row. For example, if one person viewed both xyz:view:idcard and xyz:ss:view:idcard, they would be counted once in each row. This is why the table shows 200,000 for the first page and 100,000 for the second. Adding those together gives 300,000, but that does not represent the true combined unique visitors, because the same person may be included in both numbers.

In a segment, Adobe first applies the segment criteria and then calculates the unique visitors across all the included hits. So if a person viewed both versions of the page, they will only be counted once in the segment results. That is why the segment total is much lower than the sum you see in the freeform table.

The scope of the segment is also important. A visitor-level segment can pull in all activity for that visitor, which can sometimes inflate or distort results. For the scenario you describe, creating the segment at the hit level is generally the most accurate option. This ensures only the specific page view hits that match your criteria are included, and then unique visitors are calculated from that set.

In short, the difference is not an error. Freeform tables show row-level unique visitors, which are not additive, while segments give you a deduplicated count across both conditions.

[DeebyCF]

Thank you so much!