Solved

# Per user averages

• 2 replies
• 42 views

• New Member

Dear community,

Is there a way to view the average of a ‘formula’ event over the users. Specifically, I have a parameter ‘like_ratio’ that is generated as like / (like + dislike).

I am interested in the average like_ratio per user - not weighted by likes and dislikes, but just a clean average.

If I make a segmentation chart of totals_like / (totals_like + totals_dislike) I get a chart weighted by activity (the most active user will weigh more in the result).

Is there a way to create the average of the per user like_ratio?

I have tried the avg(like) / (avg(like) + avg(dislike)) but I get a wildly different number from what I get when exporting user data and taking the avg(like / (like + dislike))

Thank you!

icon

Best answer by Matrak 23 November 2022, 16:48

View original

### 2 replies

Userlevel 1
+1

Hello !

Thus, to calculate the average like_ratio per user, I would first try computing a formula using:

• TOTALS_like / (TOTALS_like + TOTALS_dislike): this formula allows you to get your “like_ratio”
• AVG(): allows you to get the average event triggered by user

The formula would be “AVG(like_ratio formula)”. However, it does not seems to work as we cannot use AVG on another formula.

Then, I would try to add an extra event to determine the amount of unique users (using the most common event occuring such as a view page if you are using a website). You can now compute the following formula: like_ratio / UNIQUES(common event).

This should look like that:

(TOTALS_like / (TOTALS_like + TOTALS_dislike))/UNIQUES(view page)

In my opinion, this formula should get the average you are looking for.
Hope it helps!

Userlevel 3
+5

@Ingmar Oemig Thanks for your post! Matrak, provided a suggestion for how to calculate your average ‘like_ratio’  per user:

This should look like that:

(TOTALS_like / (TOTALS_like + TOTALS_dislike))/UNIQUES(view page)

At this time you cannot apply the AVG() formula on top of another formula calculation, therefore the above suggestion is the closest option available for calculating the average like_ratio per user. If this does not suffice, I am also happy to file a feature request on your behalf.

I hope this helps!

-Jacenda