Posted by PAW Notary Services on 8/10/04 1:06pm
Msg #5845

Answers to questions submitted to Paul ...

Hugh/TX: Many times there are web docs (e.g. graphics such as gif or jpg files) that are "hot linked" so they would be download from your site. That counts as a hit, but not a page view. I don't know if that is happening on your site or not.

Brenda/TX: Ask your web host if it is possible to get the raw data showing ALL referrals. You may wish to "threaten" them with moving to another host that provides it with their "cheaper" package. (My site is hosted by RCThost.com and they provide all the raw files, plus formatted views. But it doesn't mean much to me. See below for reason.)

Steve/FL: Always keep a good, clean, new toner cartridge handy for when problems arise. Whenever a company wants to dispute warranty claims they like to "point their fingers" at non-HP parts and consumables. But, just to let you know, the toner is a consumable item and doesn't fall under the warranty or conditions. However, the cartridge, though considered a consumable, is a mechanical component, but doesn't void the warranty if a non-HP cartridge is used. To avoid all the hassle and headaches, just put in the good, clean, new cartridge to see if the problem remains. Of course, there are things that have nothing to do with the cartridge that can go wrong, but it is always advisable to be able to return the printer to it's OEM condition.

-----

Why I don't view my statistics: Since I also host the Notary Public / Signing Agent Web Ring homepage on my site ( http://www.pawnotary.com/nsaring/ ), which also is the host for the ring graphic, my "hit" logs fill up very quickly with hits because of the downloading and displaying of the ring logo graphic.

Reply by HisHughness on 8/10/04 1:11pm
Msg #5846

I'm not talking about hits. I'm talking about visitors and page views. On top of that, there are no downloadable items on my web site. I wouldn't even know how to install them.

Reply by PAW Notary Services on 8/10/04 1:27pm
Msg #5853

Maybe the following description (from Opentracker) will answer your question(s):

Why do tracking services show a lower number of visitors than statistics recorded by log analyzers? The answer lies in the difference between unique visitor tracking and log analyzing. Log analyzers record all measurable activity, whereas tracking services distinguish between human activity and server activity.

Tracking service stats will show lower numbers than log analyzer stats. This is not because tracking services record fewer visitors. The reason is that tracking services are stricter in their definitions of a visitor. A tracking service should do its best to ensure that no visitor is recorded twice, and that only human clicks are counted as visits.

The reason that tracking services will report lower traffic numbers than log files is because good tracking services do not recognize the following factors as unique visits or human events:

· repeat unique visitors (after 24 hours)
· hits
· robot and spider traffic
· rotating IP numbers (i.e. AOL)

Equally important is the ability to distinguish how many unique visitors are visiting from either:

· the same ISP (Earthlink, At&t, Comcast, Cox, etc.)
· corporate firewalls, large organizations (Microsoft, IBM, Apple, etc.)

Otherwise all these users will be counted as the same visitor. This is a differentiation which can only be made by tracking cookies.

Where possible, tracking systems should only measure human events.


For years now, the standard measurement of website traffic on the internet has been ‘hits’. Hits are not a reliable indicator of website traffic. A hit is a single request from a browser to a server. When a visitor looks at a single page, many hits can be generated, both for the request itself, and for each component of the page.

Opentracker measures page views, not hits

Opentracker tracks page views. A page view is a single human event. A page view is also known as an impression. Each impression, or page view, represents an actual person who has viewed a specific web page. In this way, Opentracker differentiates between human events, and server-browser dialogues.

Opentracker specializes in human events and visitor behavior.

Opentracker tracks visitors over the long-term, and has the ability to recognize if a visitor has been at a site before. Opentracker uses browser cookies to track unique visitors over long periods of time. Examining a unique visitor’s clickstream, for example, can tell you how quickly new users adjust to site layout.

Drawbacks: we can miss visitors, in the event that a visitor clicks too quickly, i.e. does not wait for a page to load.

Case study:
Example of a discrepancy between Opentracker and log analyzer
Log analyzers do not distinguish between humans, and spiders or bots. Spiders and bots are the devices sent out by search engines to scour and document all pages on the internet. This means that a log analyzer might record an extra several hundred visits for a given period, depending on the popularity of your site. The more popular that your site is, the more often it will be visited by search engine robots. This is especially true if your content is frequently updated.

Reply by HisHughness on 8/10/04 2:55pm
Msg #5873

PAW responded:

***[Interminable technical response.]***

Thanks, Paul. I knew what I was doing when I hired you, that's for sure. Is this a great forum, or what?

So, I guess the negative difference between the number of visitors and the number of pages viewed would be one way, albeit quite rough, of knowing how frequently search engines are taking a look at a site?

Reply by BrendaTX on 8/10/04 1:22pm
Msg #5850

"Brenda/TX: Ask your web host if it is possible to get the raw data showing ALL referrals. You may wish to "threaten" them with moving to another host that provides it with their "cheaper" package. (My site is hosted by RCThost.com and they provide all the raw files, plus formatted views. But it doesn't mean much to me. See below for reason.)"

Paul, That's a very good idea.
I believe I'll smoke a 15-cent cigar.

Thanks, Brenda