Has your website been swatted by the big paw of Google's new Panda search algorithm? Mark Underwood tells SEO specialists to pay heed to Google's "suggestions" or pay the price.
Any IT professional who's spent time supporting websites has, at one time or another, been asked to help improve a site's visibility. Some of the best-conceived websites around simply aren't discoverable without some attention to SEO.
The situation periodically degenerates into a cat and mouse game, with tricksters ("black hat" search optimizers) trying to second-guess Google's techniques, and Google responding by trying to defeat those tricks. The process, as Google's core ranking team of Amit Singhal and Matt Cutts write in their blog, is "constant tuning," since its Google's Search technology is a major factor in fostering what they benevolently term "a healthy web ecosystem."
A recent, widely reported (the New York Times headlined the story "The Dirty Little Secrets of Search") incident concerned JCPenney's search rankings during the 2010 holiday season. For the common phrases "dresses," "bedding," "area rugs," and even "grommet top curtains" reporter David Segal mused, some sites other than JCPenney should have broken through, but JCPenney owned the top spot or was very near the top every time. This gaming of the Google Search system was achieved by proliferating external site references to JCPenney pages - most of them speciously connected to the subject at hand. Google calls this tactic "doorway pages," but another way to think of the web roadmap created by such links (e.g., nuclear.engineeringaddict.com), is that they are anti-semantic networks. Google made "manual" corrections to its rankings of JCPenney pages after being contacted by the Times.
In February of this year, Google made another change to its indexing algorithms for Google Search, codenamed Panda. Initially for the U.S. only, it was rolled out across all English sites globally on April 11. The ostensible objective was, as with previous tweaks, to improve the quality of search results. Google Search users have perhaps encountered sites whose content is mainly blatant copies of snips from other sites, plus sometimes unreadable lists of phrases and keywords with Please Rank Me! written all over them. Some SEO specialists have dubbed these "content farms."
According to Google, the changes made in February were comparatively dramatic, affecting around 11% of queries.
Jeopen has other tips that suggest Google is seeking more exogenous indicators of importance. These include Quora-like traffic, offline promotion, and avoiding web traffic flytraps.
Google offers "quality guidelines" on its Webmaster Central pages, and warns, somewhat ominously that "we strongly encourage you to pay very close attention." Certain practices "may lead to a site being removed entirely from the Google index or otherwise penalized." The disparaged content, produced by practices such as hidden text keyword stuffing, internally duplicated material and plagiarism (apparent or actual) are easily avoided, at least for most sites. Less obvious, but apparently affected by Panda, are paraphrased content, such as lists taken from site A and reproduced on site B, perhaps with minimal annotation.
The problems with Panda
Some have claimed that Google Panda is punishing smaller sites at the expense of bigger sites. This could be an unintended side effect of Panda's "cleansing" operations. Sites with greater viewership are more likely to have inbound links. For instance, a new site will have few inbound links, and while it serves a worthwhile purpose for a small readership, it may well be overwhelmed by large sites that, so far as Panda is concerned, appear to be more valuable.
It can be assumed that Panda attempts to compensate for this pattern by considering other lines of evidence. Cardinal Path's Kent Clark, suggests that most of the effort in SEO should be related to a site's Information Architecture. This seems more a hope than a set of algorithms at present, still it would be nice to learn that a site's place in the semantic metaverse was the result of careful knowledge engineering rather than mere popularity, or, worse, the phrasal brew concocted by hyperlink hoaxers.