1. Please advise what may be a problem with my company site
www.shrinkit-inc.com that the page jumps from position 3-5 to position
20-30 or often to position 70-80 in Google results.

How I test for Google results:
In the Google toolbar I type the same 3 words "shrink wrap film" in
order to see how our page works in terms of the results by Google.
Basically, every second or third day my page will show as stated
above.

If it shows in the higher position 3-5, after I click on 'cached', I
see the cached version that is our page current version, redesigned
some 60 days ago, with more text under picture/image links.

If it shows in the lower position, after I click on 'cached' I see the
prior, older version of the page, about over 60 days old.

My feeling is that the revised current design cached by Google is not
remaining as cached, or is being dropped, and instead the older
version pops up way down in the results. Google is re-indexing once a
month, thus after 60 days of the newer page existence the older page
should disappear from Google archived results.

Please advise what to do so that I maintain the current version as
cached and the higher position.

Background information:
- Our basic web page is www.shrinkit-inc.com presents all product
lines, and was first published in 1998.
- Our secondary web page under domain name www.shrink-wrap-film.com
was published in 2001 and focuses on presenting stream videos on
shrink wrap applications. Could this be a case that Google sees the
second page as a duplicate, which by our standard is not a duplication
despite the fact that the pages are owned by same company, however the
focus of the content is quite different
- In order to protect our trade marks we also own some other related
domain names such as for example
www.SHRINKIT.NET  or www.SHRINKIT-FILM.COM and few more, which are not
registered in directories and are not published, and are "parked" at
our provider and re-directed to the basic page
http://www.shrinkit-inc.com/ Could this be a duplication problem for
Google?

2. Moreover the page www.shrinkit-inc.com is registered and ranked in
the Open directory under:
Category: Business > Industrial Goods and Services > Packaging >
Flexible
If clicked from within the directory the page comes out as not ranked,
despite being ranked in the directory. Could this be because the page
address in the directory is shown as: http://shrinkit-inc.com/ without
the "www" portion in the address? When the address is typed with the
"www" portion as http://www.shrinkit-inc.com/ the page comes out as
ranked. This case of ranked/not ranked is a big puzzle for me. Could
this be related to my question # 1

I do hope that you will be able to explain what to do to fix the
"jump" problem and ranked/not ranked problem.

Hi there,

It could just be a glitch, with no rational explanation. if this was
occuring during an update, I would answer like this:


There is an underlying reason for inconsistent Google results, and it
is further complicated by recent changes in the way Google updates
their index.

Google is "powered by the world's largest commercial Linux cluster
(more than 10,000 servers)"
://www.google.com/press/highlights.html

In the past, roughly once per month Google would take all the data
from crawling the web and publish it as a new index. The publishing
process of uploading it to 10,000 servers usually took several days,
and it became affectionately known as the "Google Dance", because
search results would vary depending on whether the server returning
the results had been updated or not.

I say in the past because in the last few months there has been less
of a singular update, and more of a rolling update. Results have been
less static and consistent than in the past. ut in general there is
still a once a month update:

Past Updates
http://www.webmasterworld.com/forum3/2657.htm

Would somebody please summarize update Esmeralda
http://www.webmasterworld.com/forum3/14867.htm



This is further complicated by Google having a whole bunch of test
indexes which they use for, well, testing purposes. Some search engine
experts have found the URLs of the test index servers (aka
DataCenters), and they can be searched at sites such as
http://google-dance.com (select "All 9 Datacenters"). Because they
test new indexes before making them available to the public, changes
to the test servers usually indicate a fresh, new index being
published.

Which makes answering queries like yours a little bit difficult. I
searched for your keywords on every datacenter just now, and they all
have you at 14th place. I was hoping to demonstrate to you different
results on different datacenters, but right now they are all in synch.
This doesn't mean they will be tomorrow or were in synch yesterday.

Cached copies from different dates are evidence of old and new indexes
being used for searches.

Regarding your second question, I am getting a PR4 for your site, with
or without the www prefix. Any problems in this area are because
Google treats sub-domains as seperate entities, until it realises that
the www sub-domain has the same content as the domain proper. It is
best if all links to your site either have the www, or don't, but in
the long-term Google seems to work it out okay.


robertskelton-ga
Google Answers Researcher

Hi robertskelton-ga

I am glad that you have given a comment. After I submitted my Q. I
found your reply to the 15 of May 2003 20:14 PDT your eply to the
subject Google Spider in which you explained the test runs of
Freshboot with a date, that is when my page moves up to a high
position when the fresh date is shown. This is exactly what happens,
and now I understand why every second day I am way up, and other days
I am way down.

However, it beats me what could be the difference between the
freshbott and regular google spider thant puts my page as #3  or #
70+, and it has been like that for over 60 days.
Best to you
Andrew

The Freshbot isn't involved in assigning PageRank. When it find a new
site, it puts it in the search results because it can, and gives it an
arbitary PR, which in most cases is higher than the one your site
would end up with. Which is why brand new sites often do very well on
the first day they appear, and then a month later when they re-appear,
not so well.

I hadn't considered it to be a FreshBot problem because usually it is
a one-off when it first discovers a site. I have not seen this up and
down effect before. Must be a glitch.

There is no reason for the FreshBot to keep visiting your site - the
content isn't changing. But it's probably a blessing rather than a
problem - half the time you are getting a higher ranking than you
normally would.