Welcome to the Q&A Forum

I suggest using Google Analytics Experiments, with multi-armed bandit. Doing so you can set the significance level you want and the experiment will stop when that value is reached. By default the p-value there is 95%, I usually left it set to 95%. And I usually repeat the experiment a second time to confirm results.

I have likewise noticed this on several sites - that do very well in Google. Some sites were near or at the top of Yahoo page one have fallen suddenly &  drastically! Specifically, two smaller (what I call) "magnet sites" or "satellite sites", tuned to narrow band topics, and linked to the main site which also contained those topics. They were both number one, page one - and beat the main, large site on their topic. The net result is the two magnet sites fell back but Yahoo made the "mother-ship" number one in both categories. They "givith and taketh away". I too noticed a large increase in ad-space real estate. Since there's so much less room for organic results, they are obviously quashing multiple SERPS from the same outfits. I have noticed a similar result in Google years back - a satellite site will virtually disappear, but the main site goes to number one.