Re: Major Difference in stats??!!

There's a saying my father used to use .....

"Give a man a watch and he knows what time it is.   Give a man two watches 
and he's never sure."


It seems to me the vast majority of traffic stats aren't absolutely accurate 
anyway (* see below for explanation of that remark) so what counts is the 
change from one period to another.  IN other words, if you do some kind of 
promotion and your hits jump from 1000 a week to 1100, that's the same as 
seeing your unique visitors jump from 300 to 330.  It's really the 
percentage change that's reliable.  My advice is to chose one set of numbers 
and work from them.   Don't worry about discrepancies between one analysis 
and another.

(*why stats aren't absolutely accurate:

Firstly, a lot of them measure "HITS".  Great if you're measuring stats 
through your server, but if you're not a server owner,  like the majority of 
us, you want to know how many PEOPLE came into your web site, not how many 
files were called up.   There's the problem.

If a person looks at your front page, and then 2 minutes later looks at 
another page, its fair to say that's the same visit.  They're moving around 
your site.

If they look at your front page, and then don't look at any other page until 
the next day, its fair to say that's a different visit.  They came to the 
front page, went away and came back to the page 2 the next day.

But what if they look at the front page, then look at another page in half 
an hour?  Is that the same visit or not? Did they look at your front page 
and admire its artistry and sheer elegance of layout and colour balance for 
half an hour? Or did they think its all a crock of doodoo and go off 
somewhere else in the intervening 30 minutes.

Of course you can't know.  So you have to make some assumptions.  And that's 
one way the analysis programs can differ.

Also, are you counting  'unique visits' or page calls?  So if I come to your 
site, look at the front page, go to your support section, then come back to 
the front page, and go to the downloads section, the font page has 
registered two visits from me.   But is it two visits to the site?  Again, 
you have to make some assumptions.  You dont know for sure.

So my prediction is, if you try to button down exactly how many people 
walked through the front door of your on-line store, and where they went and 
how many things they picked up and browsed and put down, you are going to 
hurt your brain.  Possibly damage it and all.

If instead you take the figures from your analysis program as a GUIDE and 
concentrate on the trends instead then you'll get the information  you need 
(i.e. are people responding favourably to what I'm doing or not) without the 
aggravation and potential brain damage.

Incidentally if you buy banner advertising space based on traffic, be very 
careful of hit numbers.  Insist on visits or page calls or calls to your 
site for the graphic as the measure.   Because I can increase the hits to a 
page by simply adding more files and getting the same traffic.

Hope this helps.

Cheers,
Mike Kear
AFP Web Development
Windsor, NSW, Australia

>From: "Sunil Bahl" <sunilbahl(at)hotmail.com>
>To: hwg-techniques(at)hwg.org
>Subject: Major Difference in stats??!!
>Date: Sun, 23 Jan 2000 20:51:38 PST
>
>Dear Friends,
>
>Since the time I hosted my site on a private server (last month) I have 
>been comparing the stats provided by them (Generated by Webalizer Version 
>1.30)and the stats provided by www.hitometer.com
>
>Hitometer is saying that I have got 580 visitors since the day I hosted my 
>site. The private server stats (Generated by Webalizer Version 1.30) are 
>showing that this month it�s self I have got 572 visitors.
>
>How come there is such a large difference in the two statistics?????? The 
>difference is almost double. I do not know which one to believe.
>

>

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

HWG hwg-techniques mailing list archives, maintained by Webmasters @ IWA