Custom event’s are on the verge of being extremely useful but are held back by the lack of deletion and organization options. Currently, custom events don’t have an option to get deleted, which in theory does not matter as adding in custom events should only occur at the very end of the development cycle. But in practice it is a very common usecase. Funnels dont have this problem as setting a new funnel id makes it so the new funnel data outnumbers the old funnel data but if I set a new custom event key, it becomes unusable as I elaborate on below.
M game I released had balancing problems so I setup different beta tests and AB tests for how long it takes to achieve different levels or currency goals. However, when I opened up the analytics page it was extremely cluttered and borderline unusable in its display of the UI as you can see in the screenshot below. On top of that, there is also custom event data from out of date testing sessions as well.
This means if your game balancing or other usecases is not perfect upon the setup of the events and you need to try different event setups, the view is untenable and can become unusable. A solution would be to have an easier and more organized view to view separate A/B tests of custom events as well as an easy way to delete specific analytic entries and analytics keys.
Hi:
I wonder if you notice that in custom event we support custom field?
I saw you put all the information in custom event name.
have you considered just put your event name like TimeToPurchaseItem in the event name.
And then your item name in custom field 1 and your release version in custom field 2?
This will allow you easily see the charts with all data together.
Besides, you can filter by certain fields and even breakdown to let all item shown in the same chart for a better view?
That is a good start but I still want to see a breakdown by OS and other factors. I.E. seeing if mobile players have a higher failure rate of a custom event then PC players.
To add onto this, this is VERY limiting with only 3. It also frankly is a different problem as the docs use it for differing cohorts not data management and UX. For example, we had 30 beta tests over the course of 4 months of development work, not to mention the actual liveops. are you sure that is a good solution for the problem? Or am I misunderstanding the use of it
Sorry about that! Read over the docs and I do think this is a good solution to start with. 3 is very limiting still but a much better start then I originally though. I would advocate for better viewing UX though.
That would help a lot actually. Just a better way to ignore irrelevant UX data would be very good overall. Y’all are doing a bang up job with the rest of it though! <3
Could you elaborate more about what kind better UX you are asking for?
Do you mean the event name drop down or you mean the custom fields filtering?
I guess the event name UX issue is due to the number of event names you have. We will circle back to make it better.
Thanks for the suggestion. We will circle back on this ask.
Meanwhile I want to know what will be the issue if you just have the player Id in it? because we aggregate the result, player id is actually not a dimension. So will that still be able to show the value like what you described correctly?
From my understanding the current behavior requires a Player instance, however if a future change would require a UserId instead i’d just use my UserId for custom events
