On average, how many web sites exist on the www?  How many new ones
are created each day and how many disappear forever on an average day?
 Note, I'm interested in the number of site, not the number of pages.

This is not an answer, just some info that I came up with during my
research

Number of websites on the internet: 550,000,000,000 (Reuters)
Number of Websites found by Google: 1,346,966,000
Source: http://www.fotf.ca/if/04-05.01/articles/gofigure.html

December 2001 - 36,276,252
December 2000 - 25,675,581
December 1999 - 9,560,866
December 1998 - 3,689,227
December 1997 - 1,681,868
December 1996 - 603,367
August 1995 - 18,957
Sourch: http://news.bbc.co.uk/1/hi/sci/tech/1738496.stm

A I found this and other information provided on the page fascinating
so am including it for other people who might be interested
interested:
A source from Berkeley says that: "At 7.3 million new pages added
every day, the rate of growth is [taking an average estimate] 0.1
terabytes of new information [HTML-included] per day."
Source: http://www.sims.berkeley.edu/research/projects/how-much-info/internet.html

Search Strategy:

"number of websites"
"number of websites on the internet"

Hello there!

Just wanted to comment on the two FOTC figures cited in tisme's
comment: Reuters might have it right (though I'm inclined to think
not), but Google's figure surely relates to the number of pages/
documents it indexes, not the number of sites.  Just look at Google's
front page!  Apart from anything else, there is a huge discrepancy -
between the number of sites in December 2001 (per the BBC report) and
the number of sites reported by FOTC (date not given?).  A huge
discrepancy or a phenomenal growth in the number of sites.

I am inclined to think, however, that the Reuters figure also relates
to pages, not sites - the 550 : 1.3 ratio is striking, and is in
accord with the Bright Planet study / white paper : The Deep Web:
Surfacing Hidden Value (Michael K. Bergman) at
<http://www.brightplanet.com/deepcontent/tutorials/DeepWeb/index.asp>

The number of web sites? That's another matter - back to one of your
original questions, halejrb.  That part of the question may have been
answered in an earlier Google Answer ID: 147034 at
<http://answers.google.com/answers/main?cmd=threadview&id=147034>

Which just leaves your questions re rate of growth and
disappearance...

Good luck,  read2live

I personally doubt you will ever get any two groups to agree on this,
or find an accurate way to estimate.

The issue becomes, what is a 'website' - a collection of webpages, but
where do you draw the line between one and the other.

Are ://www.google.com/ and http://answers.google.com/ the same
website? What about http://images.google.com/
http://groups.google.com/ http://and www.google.com/ - they are all
different URLs, but all the same site really, they just activate
different options.

Then theres things like this http://www.secret-passage.com/ and
http://www.secret-passage.com/mckay/ - same author, but one has very
different content, and is really a site within a site.

http://www.bhphoto.com/ http://www.bhvideo.com/
http://www.bhphotovideo.com/ - One site, or three?

http://www.geocities.com/ - one site, or thousands?

I have yet to find a definition of 'website' that seems suitable to
base a count on. Even webpage is trouble enough, when you start to
consider dynamic pages and the like - is index.php or index.asp one
page? Or the thousands that it can become when given the right
options?

Our obsession with counting things doesn't really translate well to
the web unfortunately.

Regards,
sycophant-ga

When I posed this question, I assumed that "home" pages could be
counted as opposed to webpages.  I assumed that all websites have a
page called "default" or "index" or something that designates them as
the Home page and that there is only one Home page per website.  But
if it's not possible to distinguish a Home page from any other web
page then it's probably impossible to determine how many sites there
are as opposed to how many pages there are.

I read the Bright Planet paper on the Deep Web.  It was very
interesting.  The issues in this paper relate to my own website:  The
Multiverse Database. (www.multiverse-db.com). My website is a database
of science fiction planets.  However, you can't find my entries for
various planets by doing a search on Google.  That's because the
appropriate webpage is only generated when you run the search command
for the database.  To solve this problem I had to create a separate
index for the database in HTML.

Hi halejrb,

This is an interesting question, considering that Congress recently
approved the Library of Congress's Digital Preservation Plan:
http://www.loc.gov/today/pr/2003/03-022.html

Library of Congress's Digital Preservation web site:
http://www.digitalpreservation.gov/ndiipp/

The following information comes from a report titled "Archiving the
World Wide Web," written by Peter Lyman, and available at the Digital
Preservation web site here:
http://www.digitalpreservation.gov/ndiipp/repor/repor_back_web.html

"The Web is growing quickly, adding more than 7 million pages daily.
At the same time, it is continuously disappearing. The average life
span of a Web page is only 44 days, and 44 percent of the Web sites
found in 1998 could not be found in 1999."

That report obtains its data from this spreadsheet, which contains a
lot of information that you may find useful:
http://www.sims.berkeley.edu/research/projects/how-much-info/internet/rawdata.xls


I hope this helps.


Search strategy:
keywords: library of congress digital preservation
Google results: 
://www.google.com/search?sourceid=navclient&q=library+of+congress+digital+preservation


Regards,
sldreamer