I’m posting in a separate topic the discussion we started in the release note post for Looker 6.16 about an alerting use case.
Thanks @bens for your observation but in fact, I was not using any pivoted dimension in my data table. I made further test this morning and I noticed that if I replaced “Session Hour of Day” by “Session Hour” the alerting modal was working better with just one serie available instead of 24!! Seems that “Hour of Day” is not recognized as a time interval.
But now, I would like to see how can I achieve my use case : ideally, I would like an alert to be triggered if my conversion rate drops below a certain threshold. Let’s say 50%.
I have a doubt about “Changes by, increases or decrease by” : should we enter an absolute number or a variation rate? In my case, I have a conversion rate in the range of 0.5% so should I enter “Decreases by 50”?
Finally, I assume it’s not yet possible in the current beta version of the Alerting, but it would be great to enter another measure as the value for triggering the alert. In my case, I have several filtered meaures setup in LookML like “Conversion Rate Today”, “Conversion Rate Yesterday”, “Conversion Rate Last 7 days” everything plotted by hour/minute and could be great to say “Trigger an alert if my conversion rate today is 30% lower than my conversion rate Last 7 days”. But things for the future
Thanks a lot for your help,