fred9989 - 10:00 am on Apr 26, 2012 (gmt 0)
Many interesting points here, but a great place to start assessing the results from Google, a place which takes away some of the personal colouring I bring to my own assessment, is the Danny Sullivan page where he compares results from Google and Bing for a number of searches - it's well worth a look (follow the link in reseller's post 4445589 above, then follow the link and the bottom of the page to his article "Did Google's search results get better or worse?")
What seems clear from DS's research is that there is little to choose between Google and Bing now, at least for the "test" searches he selected, which admittedly were rather broad.
Of course what this implies about the absolute standards of these two search engines is open to question, but he does make the observation that he thinks Google has not achieved its objectives.
On a personal level, I have an uncomfortable feeling of familiarity. These results, from the health field, look so much like stuff I've seen before and would instinctively call low quality, a regression to blunders in past updates:
1) Single pages of websites ranking for important searches, not whole authority websites
2) Rankings of a lot of out of date stuff, eg blogs not updated since 2007 - 2009 (the latter blog having had no new posts for 3 years or more but it has had new links added to it from a link buying scheme as recently as Nov 2011 - not necessarily giving traction, I know, but still...)
3) Age old articles ranking
4) Age old directory entries
5) Low quality, trashy Yahoo answers ranking
6) Big branded sites ranking, but often of the superficial "magazine" type, with glib entries that have no depth or authority
Most of all, I have a clear sense of a lack of substance - an unquantifiable feeling when I look at the results that basically there is nothing I would trust or respect....though there are a couple of big name medical sites (not that they offer any products or solutions - they offer advice, rather superficial advice IMJ) and Wikipedia features too - but that's the same issue - it's information, not solutions, presented in response to a clear request for a way to solve a health problem.
And while a lot of this stuff could be the subject of a vigorous debate about how much it is worth, what really sticks out is that the only two sites on the first ten results where you can buy a product are ultra-spammy, both on page and off page, with 1000s of links from - in particular - a Chinese government site!
These links are divided into about ten different anchor texts which cover the whole range of related search phrases for this particluar query, all of which are stuffed throughout the site in abundance.
That's a clear miss for Google - but I hope to see the site drop sometime soon..... (faint hope).
And so - what has changed in the algo? No doubt many things have but my sense is strong around:
1) linked websites - in networks with a clear footprint - got demoted (ho hum!), not just in their own right but also in their ability to pass PR
2) a narrowing of the acceptable frequency of anchor text which would cause a target URL to feature in the search results for a particular phrase; too low - no chance of appearing, unsurprisingly, but go a smidgen over the acceptable upper limit (IMJ around 40%) - and the website plummets for that and all other searches
3) great sensitivity to link churn - changing, adding, or removing links too quickly (and that may be not very quickly at all) causes a penalty
4) I'm not so sure that thin content has really had much to do with this....
5) as always, though, massive link span can still keep you there, till it's detected....
Other suggestions welcome.