Actually I might have misunderstood. :) Your idea seems to not be about filtering, but adding design event in the Cohort Selector (like revenue). But that would still require some filtering on the design event (to specify what event). I will review with our teams internally.

Hey Nikolaj Ahlberg-Pedersen . I don't care about your tech solution. Event filter based on cohort is my no.1 reason to use GA...

Nano Toybox: Thanks for the feedback.

All technology stacks have pros/cons with certain functionality being more difficult/easy to achieve. Some things take longer for us.

We added install cohort filtering recently (global filter). This might solve some of your needs. In Explore Tool you can select a specific design event query and filter on install day/weeks (see image).

e.g. selecting a specific week will show metrics based on design event data sent from users who installed during that week.

We also added an "Engagement" feature where it is possible to filter to include on users who sent a specific event in the period.

As mentioned.. we are working on new functionality to support a Cohort feature with advanced options for filtering and more. This is more difficult and takes more time.

If you can, then please describe what you expect when you filter a cohort feature by an event/filter. On the install cohort users? or on the returning users? We can use this in our development. Thanks.

Hey Nikolaj Ahlberg-Pedersen. Thanks for the quick reply. I basically want the same "select metric - progression/design events; filter - certain events" in legacy GA, then check count.

For example, I want to see how many level complete events player finished each day. So I can know on average how long it takes for player to finish all existing levels. In the legacy GA, I can find that quickly. In the picture you can see the number of installs per day, the cohort data of level complete events. Here I can see that players finished 10894/1511 = 7.2 levels on D0.

I tried the explore - cohort filtering feature you mentioned. Choose 0 day since install but he result is not the same with same game and event. Not sure which one I should choose. And here I can't see the install number so still hard to use(I can search in another page and do some manual work, but what's the point of using GA if I need to do manual work?). The experience of checking this data is bad and too complicated for the number 1 data we are checking everyday. So I can't expect all my team members to do it everyday. So I won't use this in our team.

Nano Toybox Hi again..

In the above you are using only "Days since install"

You need to also use "Instal Cohort" filter to replicate.

I just tested this in Explore Tool and I could replicate the numbers using our Demo game.

I selected 3 specific days in date picker. And the same days in the "Install Cohort (days)" filter. Then "Days since install" = 0 filter.

First I selected no specific events and "unique users".

This matched the numbers reported in the old Cohorts under players.

Then I changed to "Progression event" without any filters and selected "count". This also matched the old legacy cohort numbers on Day0.

See image.

Perhaps you could provide your game id for me to test? It's the number in the URL when selecting game in our tool.

Nano Toybox Thank you for the feedback on most used metric/insights. We will discuss in the team how to achieve more easily.