AndyD wrote:

3.In my ten competitors I have ehow and wikipedia. Due to their PR, backlinks and age I can't yet compete with them but they affect my recommendations, such as number of words in the text body. Against one keyword there is an average of 2200 words. And for all the keywords it is between 1600 and the 2200 words, which is too much for a retail site. Therefore it would be an advantage to remove sites such as wikipedia from the calculations to check the real competitors.
Andy

Hello Andy,
Me again. #3 above makes sense to me if your other competitors are ranking on the first page for your keywords.
I doing an Insurance Broker client and local searches are dominated for the first 3 to 4 organic positions, but then his local competition ranks in organic listings in positions 4, 5 & 6 then a couple of national Insurance Online Quote sites that have thousands of high PR back links.
So I have only done competitive research on his local competitors and it is starting to pay off. I have his landing page for the keyword search showing in 4th position on page one of Google.
If I tried to do what it will take to compete with the big boys, (Nation Wide, Progressive, SafeCo etc. or the national mega quote sites) it would take for more time and money than I am budgeted for this project.
So IF what you are talking about is local organic listings, it has been my limited experience not to try to compete with the old big boys. At least not if there is some local competitors on the first page of the SERPs.