This reinforces my belief that online advertising produces a lot of objective data ("how many times was my ad viewed? clicked?") but benefits from not being able to tie that to outcomes companies are actually interested in ("are the ads expanding business?").
A number of years ago I read an analysis on how some large social media site had changed the order of a few important buttons out of the blue. This was likely from A/B testing showing increased engagement, but it was probably just confused users clicking on it. I bet similar things happen all the time in ads, possibly inadvertently. If an A/B change shows increased ad clicks, it's unlikely not to be adopted, even if it's not intentional clicks.
For context, this is leading to my AC unit. While hanging a light above my workbench, I noticed daylight coming in from the wall where there shouldn't be any. It appears a previous owner had pulled back the insulation and forgot to put it back - shudder to think how much money that's costed me over the last two years. Would like a hardier seal than insulation to stop water and mice, but not sure what is required.