Problems with Logfile Analysis

whats the difference between "top referring sites" a. " top search engines"

[Duumvir]

hello,
i use webtrends log analyzer.
my logfile for april shows me google with 8.000 visits at the "top referring sites".
at "top search engines" google comes with 2.400 searches for april.
in june google has only 4.000 visits at "top referring sites", but 4.600 searches (nearly twice as much as in april!) for "top search engines".
does anybody understand the exact difference between these facts?
thank you for your replys, Duumvir

[fom2001uk]

VISITS is one metric
SEARCHES is a different metric

I've never really understood WebTrends "searches" metric and I'm not sure how accurate or useful it is. I'd stick with VISITS for now.

At least you can compare VISITS between different months, and also use it for looking at percentage of overall visits.

[visca]

Top Referring Sites - ANY site that people have come from that in some way or another links or refers to your site. Ie, Bobs Home Page, 1001GreatLinks, GreatMusicHistory.com.

Top Search Engines - SEARCH ENGINES that have referred you traffic through either their directory of sites or a search result. Ie, Yahoo, Google.

This way you can get a higher level view of where your site traffic is coming from using the "Top Referring Sites" stats, while being able to zoom in and say "Which search engines are doing a good job at giving me more traffic?" with the "Top Search Engines" list.

[Duumvir]

@ visca
Thanks for your reply. But how does this fit to my observation that in one case the top referring url for google ist about 8000 a month and the top search engines ist about 2.400 for google and the other month top reffering for google is about only 4.000 and the top search engine number is about 4.600?

[Mikkel Svendsen]

One question: Do you use Google's site search? (the one served from googles servers) - remember to filter that out :)

Also, take a look at your keywords.ini file (in the Webtrends root directory) and check if all the referrering google engines have been defined. If some are missing they won't be tracked. I think there are some other threads around here regarding the keywords.ini file in case your do not know how to deal with that one ...

[Duumvir]

@ Mikkel_Svendsen (and of course all the other members! ;-)
Thanks for your reply! But we donīt use Google for site-search, but thanks
for the tip anyway.
I donīt understand the keyword.ini problem. when thereīs a certain
google engine missing i will not be tracked. so far, so good.
but how does this explain that the top reffering sites for google stay
constant for the two months and the measure "searches" is nearly twice
as high as in the comparison month? (i use a trial version of webtrends,
very new (version 7.0), so i think the keyword.ini file should be
up to date, what do you think?
best regards, Duumvir

[Mikkel Svendsen]

You should never expect the keywords.ini file to be updated - even in new versions. Many versions of the engines are not listed - make sure what you track is actually there!

If you use a reporting tool "out-of-the box" you will often run into "strange" problems like this. My advise is to varify the data you report on, the filteres and ini files you use. Unless you know what you are analysing you won't be able to debug results :)