Forum Moderators: open
Searches restricted to results from country (not available for all local Google's) looks also up, 2-5%.
Note: local engines and portals are on local language per default, so we don't have to take them into account here
I see it rising in Germany, up to 60+%...
I've just acquired FastStats - I wonder whether that prog can "pull" that information out of the logs?
lr%3Dlang_cc # for language filter
lr=&cr=countryCC # for country filter
i.e.
meta=lr%3Dlang_fr
lr=&cr=countryFR
for french language results and results from France.
restricted to local language: 12%
resrticted to local country: 3%
So much less than you're seeing.
Perhaps it's a german thing?
Just checked another site:
67% language restricted searches, 20% (!) country restricted searches.
<---goes back to see if the method is at fault...
.de
lr = 12%
.at
lr = 6%
cr = 16%
.ch
lr = 14%
Are Austrians deliberately excluding German sites (which would make some sense)? (Physical product.)
I find the most interesting idea about heini's stats is whether people are starting to use language- and especially country-only search as an exclusionary tactic to rid themselves of reams of similar generic sites in the SERPs.
That would be an explanation.
Building keyword and links maps from single user's searches would be one of the most fascinating things to look at (and I bet the top dogs do such things in beta already). However I think most people search for the same or very similar things most of the time. They are simply not interested in exploring the vast depths of the web, they want 2 things:
- Sites they are already familiar with
- The right sites
Any prefiltering cutting down on the number of results returned AND bringing results closer at home might be seen as a good thing.
The interesting thing is those extraordinarily high percentages for filtered searches I'm seeing are from informational sites as well as shopping sites.
So the idea that people tend to filter for local sites when searching for shopping sites doesn't appear to be the main factor.
Another suggestion for filter behaviour is that people use it for second or third tries, i.e. first search unrestricted, then filter down when not satisfied. My stats however tell a different story. Filtering for local results seems to be the first choice for many users, if not default setting.
Similar with filtering dependant on keywords. I see queries filtered for local results which would clearly bring up local language results anyway, as those searchterms are unique in that language.
The interesting thing is those extraordinarily high percentages for filtered searches I'm seeing are from informational sites as well as shopping sites.
There are an awful lot of generic informational sites out there these days (with more to come from the Adsense influence).
Personal story: I visited Chamonix recently and wanted to find some specific information about accommodation and resort services beforehand.
As a searcher, what was my method?
1. Try to exclude generic travel affiliate sites
2. Try to exclude non-French resort information sites
Now the argument always is that this kind of thing doesn't count as we are "experienced searchers". But Google and other search engines have been in the "public domain" for a while now and people aren't generally stupid in the long term.
As another side point, a free copy of the biggest Austrian internet magazine popped through the letterbox yesterday. One of the main articles in it: 33 Profi Such-Tricks, detailing a number of "advanced search" methods, using cache, internet archive, etc, etc. This is a mainstream magazine taken by definite non-web-professionals and illustrates, IMO, how a more sophisticated view of search is spreading out into the mainstream web public.
Heini, why don't you do a comparison of some important keyphrases that you rank equally well for in general and German-only searches, and let us know what you find? I'd do it myself but I don't have any regional sites.
To reiterate my point, statistics based on overall referrals from Google are totally meaningless.
Indeed. That was why the example I gave was taken from a site which is #1 in all Googles listed in the post - whether general, language or country is used.
(Edit added for clarification: It is effectively a 3-page site optimised for one phrase and variants.)
