I generally agree with the top comment’s notion that you should stop fiddling with budgets & targets. When we are running our evergreen program, we try to make as few adjustments as possible to those specific items, budgets & targets.
That said, you can still make adjustments to your ad schedule and leverage seasonality adjustments without sending your stuff back into learning all the time. Ad schedule changes will likely send you into some learning if the changes are dramatic, but if you’re using a large data set spanning a large window of time to make those decisions, you should still be making an overall improvement.
For me, the reasons to adjust ad schedule need to be rooted in how the business functions vs regular performance fluctuations across day of week / time of day. IE, if you have stores that are completely closed on the weekends and your customers expect a short fulfillment cycle (get it today, tomorrow, etc) it might not be a good use of money on those days. But again, depends on the business, how long your purchase / decision process is for customers, etc.