Recommendations

Make changes early in the week: If you see optimizations have improved performance, you can continue to push adjustments & optimizations in the same way.

Be conservative with optimizations: Multiple optimizations should not be made to the same data set simultaneously to allow the impact to be properly examined and determine cause and effect.  

Don’t jump to conclusions: Ideally you want a week+ worth of data to determine if your optimizations had a positive or negative effect on performance/pacing. 

Avoid making EOW optimizations