Just wanted to pose an interesting question? What happens when search engines change their relevance model?, (change happens). When you consider SEO and search engine relevance, what matters most is getting in the search index and maintaining an upward trajectory. So, before you hit the panic button when your sites rankings take on a new persona (or lose position) for your main and subsequent keywords, you may be in for a surprise if you are looking in the wrong place for the cause (the dreaded merry-go-round index).
Just like with SEO, identifying and developing the ideal information architecture, eliminating server configuration errors, building links and promoting your content does not make people purchase your products or services. The fact is, it takes the right action to produce the right results.
Much in the same way, you would think that getting in the Google search engine index would be enough to hit the top 10 and stay there, but with different data centers all crawling your content with their own unique array of algorithms to assess the criteria that comprise your pages, surviving the law of averages, chronology and authority are all part of the new Google dance.
Once upon a time, when there were less data centers and less crawling of sites, results would stay for weeks without being challenged. Now, it is possible to rise or fall 10, 20 positions or more over night, but why? What if your site is crawled and indexed from data center A in the morning, then in the afternoon data center B comes along and takes a new reading to upload to the main index, which one wins and which one yields shapes the way your pages rank in the new and improved index.
Is it merely a means of calibration, the fact that more sites are making a debut each day, or that perhaps Google is starting to shake down any site that distributes unnatural or uncharacteristic behavior (which means more penalties or possible violations)?
Now before answering that, naturally only a real Google engineer could answer that question, but from a standpoint of speculation even though I am no doctor, I would summarize the symptom as (1) a case of the rotating data center itis, known for its “merry go round” like ranking syndrome where all of the criteria being used to assess relevance score is pending the redemption ping to consolidate and rectify the data (2) a slight case of amnesia due to temporal displacement (3) new ranking algorithms being tested for relevance or (4) the elements that create the abacus like ranking factors are being shifted slightly to embrace a new parameter.
In either case, what is certain is change. Sites that held lofty positions sliding off the page and returning again in a few hours (as if nothing happened) or SERPs (search engine result pages) showing distinctly different results depending on what part of the day it is, consistently like clockwork for days on end (the 5 o’clock shadow ranking) where your site hits the top 5 for your main keyword and before you can call your mom, bring in your colleagues or show all of your friends, the ranking simply vanishes.