The Google patent application submitted in March, 2005 has generated a good deal of debate among search engine optimization experts. The patent document contains many general suggestions about the direction Google wants to move their search criteria and ranking techniques in the near future.
The document points out two areas in particular in which "there remains a need to improve the quality of results generated by search engines." (0009) These two areas are
(a) artificially inflated rank due to spamming techniques, and
(b) stale documents that rank higher than fresh ones, and therefore "degrade the search results".
These two points reveal the primary purpose of the proposals made in the March/05 patent application. In general, that purpose is to improve the quality of search engine results. The specific measures proposed in the application are meant to address the two points previously mentioned: spam which skews results inappropriately, and document staleness which results in old documents being ranked higher than newer ones.
History is more important than ever
This means Google either already gives, or intends to give the "history" of documents more significance. And not just the date when the document is created, or most recently changed. They also propose tracking the pattern of the changes in content, changes in anchor text of links, changes in numbers and quality of inbound links, changes in quality and number of outbound links, changes in other pages within the same associated group of documents.
On top of that, they propose tracking user habits and patterns over time. How users got to the page in question, how long they stayed there, how many times the particular page was clicked on when it was presented in a search...a very impressive (bewildering?) array of factors.
In fact this is an ingenious attempt to solve the "spam" and "staleness" problems at the same time. The major assumption seems to be that up-to-date "relevant" content -- the kind the search engines are supposed to be giving us -- will be regularly updated, will be inter-connected by an ever-increasing (and regularly changing) group of inbound links. In other words, links will come and go, changes will happen gradually, and "spikes" in either traffic or increased link activity will be sure signs of spamming activity.
Conclusions
Whether all of these measures will ever be implemented or not is almost irrelevant. The future has been defined, and it is up to creators of websites and online marketers to make the most of it. The most important conclusions we can take from the patent application is that the history of our pages matters. More specifically,
-- Rapid and wholesale changes in content will be looked upon with suspicion
-- Rapid increases in numbers of inbound and outbound links will trigger red flags
-- Changes in anchor text that alter or remove its relationship to on-page content will be suspect
-- Lack of regular and steady (but not radical) changes will get your pages labelled "stale"
-- Links that were valuable last year (or month?) will not be as valuable this year (or month) because they are becoming "stale".
In other words, keep adding content, keep upgrading your pages, keep improving and adding new ones, continue to get new links, and freshen up your old ones if you can. But don't do any of it too quickly.
Think of this "history" component as a method of measuring change. It may seem unreasonably vague, but in the new world order, change has three speeds: Too Slow, Too Fast, and Just Right.
Subscribe to:
Post Comments (Atom)
0 comments:
Post a Comment