> I don't want to get into some sort of 'WT is overly creative' flame war
> here and waste everyone's time, so this will be my last post.
You already have...
> Okay, maybe pure twoddle was a bit wide of the mark - let's say
> genetically modified twoddle 8). It is better to be simpler and accurate
> than to be fancy. At best, the WT reports below are wrong, at worst
> totally inaccurate and misleading. If your client came back to you and
> asked you to explain the logic (and ergo accuracy) behind these reports
> would you be totally at ease telling them? I know I wouldn't and that's
> why I don't use the software.
No these are not wrong - they are just guides. If you include them and
make out they are gospel, then YOU are wrong.
> pathways within site
> entry page
> exit page
> visit length
> page views per visitor session
> first time users
> etc etc
>
> For an explanation of why these are inaccurate
> http://www.analog.cx/docs/webworks.html - have you actually read this? I
> would hope that you're not sending out reports without having done
> so........
The same explanations are available in the WT manual, have you read that?
Both documents point out "might's, if's and but's" and these proviso's
should be passed on to the client. For instance, I personally do not
include pathways within site or entry pages as there are too many
variables to make sense of this. However, exit page and visit length can
be tightened down more. For example, if a requesting ip is not 'seen'
again for a specified period - say 60 mins, you can reasonably assume
that person has left the site. But as with all the stats (no matter how
you analyse them) there is a margin of error.
My point is simply that if the data used is from the same source then the
basic stats should be the same. Analog is not better or worse than WT in
my opinion because the resulting report data must be the same - unless
one of them is inventing data!
You are simply putting too much emphasis on the fact that WT has more
config/graphing/table tools than Analog and calling this 'wrong'. It is
the interpretation of data that is right or wrong, not the analysis
program.
This is also my last post on the matter!
Brian
------------------------------------------------------------------------
This is the analog-help mailing list. To unsubscribe from this
mailing list, send mail to [EMAIL PROTECTED]
with "unsubscribe" in the main BODY OF THE MESSAGE.
List archived at http://www.mail-archive.com/[email protected]/
------------------------------------------------------------------------
