growth numbers

From: Jim Pitkow (pitkow@parc.xerox.com)
Date: Fri, Mar 05 1999


Message-Id: <4.1.19990305115534.02890cc0@mailback.parc.xerox.com>
Date: Fri, 5 Mar 1999 11:58:13 PST
To: www-wca@w3.org
From: Jim Pitkow <pitkow@parc.xerox.com>
Subject: growth numbers


Last fall I did some calculations to get within an order of magnitude of
the growth rate of the WWW and of WWW pages.  Here's what I came up with.
Henrik suggested we put this up on the WCA page, so I'd like a reality
check.  Could you please check my reality?  OCLC, can you provide more?


The best source of data that I am aware of is from researchers at Compac
SRC, see:

	http://www.research.digital.com/SRC/whatsnew/sem.html

The reason I place more confidence in their numbers over those published by
others is that the same methodology was used for all measurements.  Other
published numbers tend to either change their methodology, are not repeated
across time, or measure the size of the web as a side effect of performing
some other task (like running a search engine).  So, Compaq SRC's numbers are:

June 1997		120 million pages
November 1997	200 million pages
March 1997		275 million pages

Growth rate: 
	= doubling in slightly less than 9 months 
	= 19.4 million pages per month
	= 16% more new pages per month
	= 7.5 new pages per second 

What I find interesting about the growth rate, is that we can compare it to
the number of new WWW sites added per month.  For this I turn to the folks
at OCLC, see:

	http://www.oclc.org/oclc/research/projects/webstats/statistics.htm

Where they find:

1998		2,851,100 Web servers
1997		1,570,000 Web servers

Growth rate:
	= 106,000 servers per month
	= 6.7% more servers per month
	= 1 new Web server every 2 seconds

which is slightly slower than the rates for the total size of the Web.
This means that average number of pages per site is increasing as well.  I
do not have any evidence of this to support this conclusion, but should
have it in a month or so from sources here at Xerox PARC.

Jim.