Over the years, as we've made various changes to our sites, I think I've seen a pattern in which our Adsense revenues (and eCPM, and clicks and CTRs) noticeably (albeit not dramatically) drop shortly after we make any significant changes to the site.
Eventually, it seems like the revenue gets back up to the pre-change levels (or a bit higher if the change had a beneficial effect on visibility of the ads, or average time on site, etc.). But, before seeing the benefit of the change, or recovering from the adverse fallout from the change, there always seems to be a period when revenues dip.
Lately I've been wondering whether this pattern is real or imagined.
I'm curious -- have others observed this same pattern, or something similar?
If so, do you have any hunches or theories about what might explain the pattern?
My best guess/favorite theory (with a bit of anthropmorthizing) is that the changes we make to a site initially tend to confuse Google -- or at least make it question what it once thought it "knew" about the site and its content. So, Google goes back to square one and starts experimenting like crazy, trying to figure out which ads it should show in the various slots on each page. Eventually it figures out the page really hasn't changed very much, and the ads that used to be ideal are still ideal (or perhaps some tiny, subtle changes are merited due to some minor change in the content or layout of the page), but in the meantime there is this long, frustrating period in which it is bouncing around doing all sorts of experiments, each of which is suboptimal, and thus revenues and eCPMs tend to decline during this "confused" period.
I'm wondering if others have had similar experiences, how long the "confused" period seemed to last, and what might be causing it.
