QUOTE(DanThies @ Feb 22 2005, 01:48 AM)
Randfish, your tool might be slower, but the resulting estimate is worth a whole lot more than KEI. KEI just combines a bunch of numbers to make another one, and it's not any more useful than the numbers you started with were by themselves. Your tool gives a pretty good SWAG at the level of effort required to get in front of searchers.
Dan I'm afraid I can't agree with this. Everyone appears to have forgotten what KEI is so I think a little Economics 101 is called for here.
The basis of most modern economic theories is the Supply/Demand ratio;
Low Supply + High Demand = Good Business Opportunity
KEI is an application of this for Keyword Research - KEI = Keyword
Effectiveness Index.
In the case of Keyword Research:
Supply = Number of Competing Results
Demand = Search Volume
So, Low Number of Competing Results + High Search Volume = Good Keyword or Phrase
KEI =
(Demand^2)/SupplyNow we have defined our terms we can apply them to some keyword research, firstly we need to
define our
Supply and
Demand values:
Supply1 = Google broad match
Demand1 = Overture Search Volume
So if we take our keyword list and apply KEI values to it and sort the list by KEI in
descending order we can say that the keywords at the top of the list will have a higher Demand
compared to Supply than the keywords at the bottom of the list.
The value (or relevance) of this information is dependent on the value (or relevance) of the Supply and Demand figures obtained.
Personally, I tend to use:
Supply2 = Google intitle:keyword phrase
Demand2 = Word Tracker Count
This gives me a KEI that, IMO, has more relevance than above.
The KEI is only as good as the data that it is computed from and is only applicable for comparing keywords from the same list - you cannot treat it as an
absolute value (i.e. "any keyword with a KEI of greater than 10 is good" is obviously untrue), I tend to focus on the keywords in the top 10%-20% of my sorted list (after I have filtered out all the rubbish).
It is very misleading to tell people to ignore KEI, you may as well tell them to ignore WT/KD/Google Supply and Demand figures and not bother to do any keyword competition research beyond finding a list of keywords - just get a bunch of related keywords and use them all indiscriminately - they all have equal value.
For an excellent explanation of KEI have a look at www.wordtracker.com/database_help/keihelp.html