We have a report that finds new threads where non-staff users have answered each other.
We originally built this report to be able to award users with badges for being helpful to other users. Over time, though, we found that it’s a good way to catch incidents in our SaaS service, when other users show up to say “I’m having this problem too”.
We used to check this report manually, multiple times a day. And then I realized I could automate it to send us a PM when the report was non-empty. Cool!
Since this is an incident-catching report, I set it to run every 30min. All good.
Until the first weekend, when somewhere around Friday night, early Saturday morning a helpful user chimed in on someone else’s thread. Within 30 min we got a message. And again 30 min later, since no one was working to take steps to clear the report. And 30 min after that. And again 30 min after that… All weekend. Everyone on my team came in to, I think it was about 125 messages.
My teammates were… unamused.
(BTW, I showed them in standup how to mass-dismiss this sub-class of notifiations. They’re still not happy with me.)
They think I should ask for a checkbox to Skip sending PM if the results are the same as last time. But that would require storing the old results, which I suspect would be more trouble than it was truly worth.
So instead I’m asking for the option to schedule this more tightly with the ability to choose multiple days of the week, and a time range per selection. E.G.
S M T W Th F Sa
x x x x
From: 6 a
To: 5 p
---
S M T W Th F Sa
x
From: 6 a
To: 3 p
