Skip to main content

Hi All, 

I’m curious to know what is the differences between UNIQUES and PROPCOUNT? In one report, I build an event segment (Any event, group by UserID). Then, I use 2 custom formula: 

(a) UNIQUES (A)

(b) PROPCOUNT(B)

At the bottom of table, there’s one line breakdown of Formula A: UserID=none. However, in the result of PROPCOUNT, noticed that “UserID=none” is excluded from calculation. Is this exclusion in PROPCOUNT formula due to “UserID=none” is not an unique value? But, how come it was included in the UNIQUES(A) breakdown.

 

Can someone enlighten me on this?

 

Thanks

 

The key here is that (none) does not mean that User Id has been set to (none), but that User Id has not been set at all. There are events where this property is not present at all.

UNIQUES looks at your event on the level of users, and always shows the full distribution of different values for the given property, including blank values, i.e. (none). You should find the same situation for other properties as well.

Meanwhile PROPCOUNT is not interested in users, only the actual values for the chosen property. Since (none) is equal to blank, i.e. there is no value available at all, those are excluded. So in this case, as you suspected, (none) is not considered a unique value, since it’s not really a value at all.

Reply


Terms & ConditionsCookie settings