I’ve seen people constantly suggesting to use recurring backend workflows rather than recurring events, even in situations in which a recurring event is a potential solution. Examples would be when a event needs to happen on a regular schedule like every day or every month.
In my attempts to find the optimal way to build in the age of WUs, I had to test the two approaches. The test lead to interesting results and some confusion related to WU consumption with recurring events.
If you use a backend workflow setup to be a recursive, you will incur WUs of 1.30 WUs…0.7 WUs is to run the action of schedule backend workflow, and then 0.6 for running the action of ‘send email’ in my simple example test, but then you get charged another 0.7 WUs to run the schedule backend workflow action again to get the recursive functionality.
So this would assume it costs, 1.3 WUs for each email sent (if you were only sending an email)
However, if you use a recurring event, the cost of WUs for setting the recurring event is around 1.22 WUs…this is 0.6 for running the action of ‘set recurring event’ and 0.62 WUs for the necessary update which consists of a read and write to the relevant item. Reading an item uses at least 0.015 WU + more for each byte, and writing uses 0.5 WU (don’t fault me for math not adding up). Then you also get charged the 0.6 WUs for the send email action.
But, everytime after that first initial setting of the recurring event, it costs only 1.24 WUs, which is the 0.6 WUs for the send email action and the 0.64 WUs for the necessary update which consists of a read and write to the relevant item. Reading an item uses at least 0.015 WU + more for each byte, and writing uses 0.5 WU.
So in summary, if you are going to use recurring events to send an email daily, it will cost after an entire 365 days a total of 453.825 WUs compared to a total of 474.5 if you use a recursive backend workflow.
Put that up against an app with only 1,000 Users and you’ve got a difference of 20,675 WUs…not a dramatic savings, but a savings and hopefully reducing overall stress on Bubble systems (which the stress on bubble systems is why we have WU pricing to begin with)…Everybody should do their part to ensure there is less stress on Bubble systems.
BTW, if you do not believe this, test it for yourself and share your results, I’d love to be proven wrong, or perhaps have other developers find other numbers related to recurring events.