Whats in Virginia?

I've just been looking at the web server stats for a new client and have
observed some strange goings on!

Basically in July the web site received 17,800 visitor sessions from the USA
of which 95% were from Virginia. This represents over 50% of the web site
traffic. Considering that the market for the web site is solely within the
UK I find this surprising.

I am assuming that this is created by a Proxy or something similar but don't
know what/who is located in Virginia. Its not obvious looking at the rest of
the information. Any ideas?

--
Martyn Fewtrell
mfewtrell (AT) networkclub (DOT) co.uk

http://www.networkclub.co.uk

Probably the AOL proxy.

so how does one collect accurate web-useage stats when the search engines
keep being our most frequent visitors ?

On Mon, 4 Aug 2003 23:50:03 +0100, "Brothermark"
<yonnermark (AT) hotmail (DOT) com> wrote:

Probably the AOL proxy.

so how does one collect accurate web-useage stats when the search engines
keep being our most frequent visitors ?

Huh?

"Brothermark" <yonnermark (AT) hotmail (DOT) com> wrote

Probably the AOL proxy.

so how does one collect accurate web-useage stats when the search engines
keep being our most frequent visitors ?

There is really no 100% accurate way to do AOL because their servers cache
your site anyway. the hits that your web logs are showing are not recording
the cached hits.

At least this is my understanding.

Does anyone else have a good solution to this issue?


James Taylor
www.AICompany.com

On Tue, 05 Aug 2003 01:45:44 GMT, "James Taylor"
<james.TayNOlor (AT) SPAaicompanMPLEASyE (DOT) com> wrote:

"Brothermark" <yonnermark (AT) hotmail (DOT) com> wrote in message
news:bgmntf$q5qoc$1 (AT) ID-132564 (DOT) news.uni-berlin.de...


Probably the AOL proxy.

so how does one collect accurate web-useage stats when the search engines
keep being our most frequent visitors ?

There is really no 100% accurate way to do AOL because their servers cache
your site anyway. the hits that your web logs are showing are not recording
the cached hits.

At least this is my understanding.

Does anyone else have a good solution to this issue?


James Taylor
www.AICompany.com

http://webmaster.info.aol.com/caching.html

"Martyn Fewtrell" <mfewtrell (AT) networkclub (DOT) co.uk> wrote

I've just been looking at the web server stats for a new client and have
observed some strange goings on!

Basically in July the web site received 17,800 visitor sessions from the
USA
of which 95% were from Virginia. This represents over 50% of the web site
traffic. Considering that the market for the web site is solely within
the
UK I find this surprising.

I am assuming that this is created by a Proxy or something similar but
don't
know what/who is located in Virginia. Its not obvious looking at the rest
of
the information. Any ideas?

AOL is based in Virginia. Most stats programs will show AOL users as
eminating from Virginia, and I suspect will show similar results for other
major ISP's (Colorado for Verio, etc.) when in actuality the user could be
anywhere.

Take all stats with a big grain of salt...

First off, yes, the AOL.COM domain is registered to Virginia, so any
and all AOL users will look like they are from VA to many webstat
programs. It has been this way for years.

As for search engines being the most frequent visitors - do you mean
REFERALS from search engines, or direct hits from spiders? If you
mean actual users being refered by search ,that is a good thing - you
just want to track which search engines are most important to you.

Josh



"Brothermark" <yonnermark (AT) hotmail (DOT) com> wrote

Probably the AOL proxy.

so how does one collect accurate web-useage stats when the search engines
keep being our most frequent visitors ?

"James Taylor" <james.TayNOlor (AT) SPAaicompanMPLEASyE (DOT) com> wrote in
news:u9HXa.12737$Ad4.5129696 (AT) news3 (DOT) news.adelphia.net:

There is really no 100% accurate way to do AOL because their servers
cache
your site anyway. the hits that your web logs are showing are not
recording
the cached hits.

At least this is my understanding.

Does anyone else have a good solution to this issue?


James Taylor
www.AICompany.com


http://webmaster.info.aol.com/caching.html


So this means I CAN control whether or not AOL caches my stuff.

Think long and hard about it though: there are penalties.

As I understand it, the reason ISPs do this (and AOL is not the only
one: most commercial ISPs and universities also set level of caching on
their users as default) is that bandwidth is expensive: for site owners
and ISPs, and loading from a local proxy is faster.

If you force every page to be loaded direct from your own server, your
bandwidth costs go up, and your websites may appear slower to some
users. For many sites, the advantages of more accurate monitoring
aren't as great as the disadvantages of having users annoyed by a slow-
loading site.

The commands can be overridden by the proxy server owners: I understand
that on the JANET network for example (uk university network) the proxy
system largely ignores them, because the bandwidth costs of observing
them would be prohibitive.

I've even come across proxy server admins that monitor what is not being
cached, and override the settings on popular sites to save on bandwidth.

The cost saving of forcing all access through a network of proxy servers
was partly what allowed the creation of 'free' ISPs in the UK, and the
resulting expansion of the internet market in that country.

Victoria
--
Clare Associates Ltd
http://www.clareassoc.co.uk/
01822 835802
--

"Brothermark" <yonnermark (AT) hotmail (DOT) com> wrote:

so how does one collect accurate web-useage stats when the search engines
keep being our most frequent visitors ?

This is an old, but relevant (imho) discourse on capturing web usage
stats from web server logs:

http://www.goldmark.org/netrants/webstats/

--gregbo
gds at best dot com

"lostinspace" <lostinspace (AT) 123-universe (DOT) com> wrote in
news:f7YXa.17399$Vx2.8847327 (AT) newssvr28 (DOT) news.prodigy.com:

If some time and thought has been taken in the application of
htaccess? Than accurate stats beyond the imagination of most
webmasters and anything beyond the comprehension of a 1995
statistician are achievable.

How do you use .htaccess to get over the problem of requests for cached
data not being reported, or of one visitor making requests from several IP
addresses?

I agree that that article is a bit one-sided, but much of what it says
still applies to most webstats. Would love to know how yours are
different?

Victoria
--
Clare Associates Ltd
http://www.clareassoc.co.uk/
01822 835802
--