Quote:
|
This just cant happen becuase search engines arent people and cant look at websites and see if its good for users, so they try and see if its good for users my picking keywords, giving weights to words, anchor texts etc etc and by the frequency of updates (dependent apon the subject if updates are needed!)
|
They don't need to analyse the page to see if it's good for users, they only need to return a document that is relevant to the current query when real people are searching.
Webmasters, SEOs etc tend not to search like "real" people do, we search like bots do and use a few keywords to see where the pages from a site we are interested in happens to rank.
The upshot of this is, the Google SERP position isn't always totally dependant on what words we put into a page, or how many times they are crammed in there, Google is steadily becoming a natural language search engine so it is more dependant on the users query. Therefore
how we put those words into the page comes into play more. Using the (key)words in a natural way including the (so called) "stop" words is going to make your page more relevant to the
way that users search.
So how you write your copy is going to depend on the target audience.
BTW. Simply scoring the page on word count and simplistic algos for weighting on embolding, proper case, headings etc went out with AltaVista (although MSN shows some vestiges of this) .
Back to the updating topic,
I have yet to see any kind of evidence that updating on a regular basis makes any difference to anything. One site I run has had no changes at all in the past 18 months because there is nothing to change, the product is the same, the very narrow target market is the same.
This however has not reduced the crawl rate, not reduced the rankings and has not made a difference to the traffic & enquiries it gets.
I'll see if any changes happen in the next few months as the product has been substantially expanded to increase the target market, so an update is on the cards. Though I do expect it to make a difference to traffic as there will be several new pages for the extra targets.
The other thing to consider is of course there is no reliable mechanism outside of the SEs own datacentres to determine when or even if a page has been updated.
The HTTP Last-Modified header can be spoofed and does not always exist in the response headers
Even the HTTP 304 (Not Modified) response is not reliable as pages created by code will always return 200