Daily Breakdown of a Metric Based on Segment: Will the segmentation happen on the daily level?

[Royyyyy]

Let’s say I have a segment defined as visitors who used X feature in the app then I segment a panel using this segment, i.e. only filtering out users used X feature. Then if I do a daily breakdown on unique visitors count. Does the daily number equal to visitors used X feature on that day (segmentation happens on daily level) or visitors did anything on that day but have ever used X feature (segmentation happens on panel level)?

Similarly, if I define a segment in a metric definition and do a daily breakdown of that metric. Will the segment I defined be applied on daily level or the entire data in the panel.

Thanks

[Alexis_Cazes_ · Accepted solution]

I am under the impression that it would be visitors did anything on that day but have ever used X feature (segmentation happens on panel level)

So basically I believe when you use a segment with visitor container where feature X was used, the segmentation will be used against the full timeframe pool of data that is selected and not the time dimension granularity that is applied.

Think of it as a box, your pool of data is a box and reporting will start by selecting this box, then inside this box you will find several visitor boxes. What segmentation would do is open each visitor box validate that it can find the feature X and then select the FULL VISITOR BOX for reporting. When you apply the granularity it will report against the collection of visitor boxes that were extracted and not only against the day worth of data box.

This is how I understand it 

[Jennifer_Dungan]

@Alexis_Cazes_  is correct... Visitor Level segments are processed first from the entire history of a user (i.e. has a user ever done this action, or visited this page, etc ever in their lifetime), then any user that fits that lifetime profile is shown in the report if they came to your site in the report period (and using a daily breakdown.... that they came to your site on that day)

If you are looking for "visitors who used X feature in the app", your best bet is to actually use a Hit Level segment...

Basically, you are looking for the use of "feature X", the actual use... you can then stack this segment with Unique Visitors metric, or Visits Metrics, or a custom event for feature X.... Adobe will use the metric in conjunction with the segment to show you visitors who performed feature X on day Y.

Example:

I have a custom metric to track Feature X, so I can create a segment at the hit level to identify its use:

Now, when I pull this into my freeform table, I can use that segment to see data in multiple ways (note that the direct "Feature X" metric is the same as a raw value, as it is nested under the segment, and also the same as occurrences when nested under the segment (so if you don't have a custom metric, but are identifying Feature X through a different method, note that occurrences can be used as a generic "all instances of" metric when nested in this manner. You will also notice that the Unique Visitors and Visits metrics when nested are less than the occurrences (as expected, as some users will be performing feature X multiple times per month, and potentially multiple times per day:

Now, if you are trying to do complex reporting where you need to know users using Feature X in a month who also do Feature Y... you could create a segment for Feature X by visitor, but add a custom date range into your report such as "last month" or "last 30 full days" or something that makes sense... sadly, dates in segments are kind of generic - they aren't really flexible to the date ranges in your report:

I did a session on using dates and time in Adobe, and I touched on Visitor Level Segments in this session:

https://events.bizzabo.com/411737/agenda/session/922401