Google released its annual report. It reads very similar to the report filed before its IPO. It is long, but it is worth at least skimming over. What strikes me most about it (as the pre-IPO report did as well) is how much of it focuses on the myriad ways Google could fail. Have you noticed that the bulk of the report details all of the possible things that could happen to cause Google to shrivel up and die? Ok, I don't normally read annual reports, so maybe all of them have the same tone of pending doom and gloom. But geez, after reading this, I could never feel comfortable investing in such a risky proposition. (Not that I ever invest in anything, mind you).
Google repeatedly brings up the issues it faces in dealing with things like server load, electricity requirements, etc. This ties in with all of the "Google is broken" posts you often see in the forums (and yes, I've made a few of those myself). Those who blindly believe that Google has no problems handling all of the hardware, software, and data volume issues, need to consider that Google probably struggles with these issues every moment of every day. And they won't always have a handle on it. All of which leads me to my latest thoughts about the last few months of search results.
Having read a large majority of all the SEO related posts over the years, this is the first period of time in which so many have centered on the following issues (in my memory). For the past several months, as we all know, the Google datacenters have been wildly out of sync with each other. Not the small differences of times past, but large, obvious differences that change many times a day. Different theories have evolved about it including different sets of algos being run either for testing purposes or to confuse SEOs, different sets of data being used, even random shuffling of results. However, combined with this has been a large amount of posts detailing old caches cropping up (more so than in the past), backlinks being updated, then reverting back, then updated again, then reverting back, and on and on. My personal observations show that different SERPs seem to relate to old data vs. new data.
I simply cannot agree that what we have recently seen is the result of Google supplying these results by design. I cannot agree that it is merely to confuse SEOs. Using old data and delivering massively different results on a continuous (moment by moment) basis, cannot possibly be a sound business decision. The only thing that makes sense to me is that Google is having some sort of data handling problems. No, I don't have any kind of proof to offer. I only have my observations (and those of many others), logic, and intuition to guide me. Nothing to do now but continue to observe.