Roy Schestowitz wrote:
>
> I want to come back, but there is so much spam here. Virtually every OP comes
> from Google Groups... :-(
I just ignore them. There are still good folks here. One thing, though,
i wrote a post on word tracker versus google trends and no one replied!
I mean nobody. That was not what i expected. Here is the post again, and
if you have any comments, i would like to hear them:
-----
Subject: trends vs. tracker
Date: Wed, 23 May 2007 09:46:49 -0700
From: catherine yronwode <cat@xxxxxxxxxxxxxx>
Organization:Herb Magic
Newsgroups: alt.internet.search-engines
Having had some success in times past with word tracker, i just dipped
into google trends -- a far more interesting site due to its graphical
interface, but also less poductive of numerics.
Plus -- it uses actual google as its resource, which partically
compensates for the total lack of numerics.
Plus -- it tracks three year trends, not just a recent snapshot.
Neutral -- the news benchmark is interesting but so far has not affected
any of my key terms. However, i did try working the news benchmark
against searches on some media terms (e.g. <britney spears> and saw its
application to those who might happen to work in that field of endeaor.
Plus -- it allows for boolian operations (e.g. <pagan -jose> to
partially isolate pagan religious content from jose pgan sports
content).
Minus -- it makes no suggestions; you have to work off your own keyword
lists (mine were generated by word tracker)
Plus -- up to 5 comparisons at a time.
Minus -- without numerics and with only 5 comparisons per query, you
have to run and re-run sets of comparisons to create a graded list of
search term popularity -- a real time-waster, that.
A tip -- establish some benchmarks of your own when working with
relative (non numeric) comparisons. I used the terms <ebay>, <help>, and
<learn> as high benchmarks, and i also had specific benchmark terms in
my field located at middle and lower levels -- your own field's
benchmark terms will vary by what you are trying to accomplish, of
course.
Neutral -- the graphical charts are relative (since there are no
numerics) and there is an appearance of artificial flattening on
lower-volume search terms when comparing a popular term against a series
of less popular terms.
A tip -- if flattening occurs, replace the highest scoring benchmark
term with a lower scoring benchmark, to see detailed trends on lower
popularity keyword terms.
Minus -- the order of the 5 terms is important -- if the first term in
the comma-delimited string is too low to score, the entire set of 5
comparisons will come up blank.
A tip -- keep a banchmark at the beginning of the string to avoid query
failures due to some terms being too low to score if they were in first
place.
A final plus -- pretty colour choices.
All in all, despite the time wasted on having to make comparisons, and
the fact that you must generate a list terms off-site becuase it will
not make suggestions, this is still an attractive feature and one well
worth using.
Any other comments, tips, or suggestions would be greatly appreciated.
cat yronwode
http://www.luckymojo.com/hoodoo.html
help youself learn love, money, protection, and lucky spells
|
|
