A Google friendly site structure
Topics: page and site building/organization principles
Google does take in great consideration well organised
sites that group related content into discrete sets. Pages that link to
lots of related content on a similar subject within your site will have
the key topic reinforced by the inclusion of the sub-pages. This
concept introduces the term "themes". It means that a
section of a site that contains a large amount of information on a
given topic will get more importance than an another site's single page
on the same topic. The site is judged as a whole, the context of each
page is taken into consideration.
The suggested site structure (see previous page) is
made for this purpose.
Over the web you can find many theories starting from the original
pagerank algorythm: PR(A) = (1-d) + d(PR(t1)/C(t1) + ... +
PR(tn)/C(tn))
BUT:
1. That's the ORIGINAL pr-algo, and it's
known that Google's team has adjusted and modified more than once, so
techniques based ONLY to this algorythm could be useful today, but not
tomorrow
2. Google sincerely states on their website that their search results
are "automatically determined by more than 100 factors, including
our PageRank algorithm."
(http://www.google.com/webmasters/4.html)
Time is better spent getting your site optimized with title tag, proper
use of headings, keyword rich content, smart site navigation and a some
good directory listings before totally dedicating your work to
PageRank.
Your home page should not contain links to ALL your site's pages for
many reasons. I.e. if you have a large website, with a large amount of
informations therefore many internal pages, homepages with a lot of
link within could be penalized as banner farm; too many outbound links
(even if they're linking to pages of the same site are outbound link!)
can seriously weaken your pagerank.
 According to Pagerank™ specs, you should not
have more than 100 links on your home page. Anyway, consider that a
good site optimization, a wise topic-division of your contents and an
easy internal navigation can reduce home page links to 10/20, including
links to RCP (see Rich Content Pages section)
Links from your home page (first level links) should
take your visitors (and Google spider, of course) to the main secondary
areas of your site (divided on a theme-based scheme). Links from your
secondary pages (second level links) should take to inner pages of the
same theme, and so on.
Once again, this structure will help keeping your site well organised
(easy to maintain), easy to be spidered and indexed, and will get the
most of the relevancy for each of the information you have published.
Remember all pages must link to at least one page on a level below AND
(if you want) to your home page.
Basically,
you should not use a frames structure unless you absolutely have to use
them. You will get the most from google spidering using
standard HTML structure because you will be able to emphasize a large
number of keywords!
Some tricks can be used when your design forces you to a frame
structure.
1. Each of your linked pages, AND menu too, MUST have a link to the
home page _TOP targeted (i.e. <a href="http://www.googlerank.com"
target="_top">)
2. Include, in menu frame, a SITE MAP, possibily with a _BLANK target.
This will ease Google to follow ALL your site link, and index it the
right way.
3. Optionally, you can use some simple javascript in the HEAD section
of each page, so orphaned pages, even when directly linked, will get
back into their parent frameset.
All website's pages at root level?
Many experts suggest to publish all pages in the
root directory of your site (same dir of your home page); this because
usually pagerank value degrades of -1 in pages included in directories
deeper than root, and keeps on degrading of -1 as you create deeper
directories.
That was true in the past. When published, new
directories' pages inherited the main page's pagerank value degrated of
1. Then, after a couple of updates, the right pagerank value was set.
Now all newly published pages start with a 0 value pagerank (because
they're not indexed yet). After a couple of fresh crawls (but
it may take longer) they get the correct pagerank, according to
inbound links quality, website's pagerank and internal linkage,
regardless of the directory.
It seems Google has corrected that little bug that, for instance, gave
new Geocities websites a pagerank value of 7 because they inherited the
main Geocities pagerank (which was 9, or 10...I don't remember,
honestly). Obviously, after the first Google dance those aggregated
websites had their pagerank downrated to 0, because the value of 7 was
NOT their REAL, calculated, pagerank.
We
still suggest a clear website's structure (tree-like), where different
topics and contents are distribuited into different directories.
This will help Googlebot when assigning a relevance to different parts
of your website; it will be easier for you to maintain your website
(also considering that the major part of your contents should be made
in Html); internal linking system will be clear to your viewers and to
Googlebot when indexing your pages.
Do not create more than three directories levels deeper than root. Your
pages must be reached within four clicks.
i.e. www.googlerank.com/ranking/fac/ebook/graph/img/cont/index.html
is WRONG.
|
|