1 reply to this topic

[scrooks]

Greetings,

This is my first post, and I'm curious about what the community feels about a concept I've been using in keyword research. Feel free to bash, suggest improvements, or compliment.

Currently, I work at a legal aid non-profit which assists low income individuals with civil legal affairs. We are launching a new website soon, which I built, and I am in the process of performing keyword research. I watched Jill's tutorial on Lynda.com regarding SEO to get me started, and it was extremely informative.

One tip above all other tips she gave caught my attention in the "Determining Competitiveness" section of her tutorial. She mentioned repeatedly that, when researching keywords in Google by performing allintitle searches, you want the global searches for a keyword to be as high as possible and the allintitle results for that keyword to be as low as possible. I was an engineering major in college, and being a big fan of math, this struck a chord. Basically what she was saying was that you want the ratio of your global searches to allintitle results to be as high as possible.

As an example, if you had 4,000 monthly searches for a term, and 2,000 allintitle results for that term, this would give you a SAT ratio (stands for Searches, All in Title) of 2. This would be a decent keyword to optimize for in your title tags.

I then discovered Google's additional advanced search operators (allinurl, allinanchor, allintext) and started doing the same thing; that is, obtaining the number of number of allinurl, allinanchor and allintext results for a particular keyword, and making ratios out of them as compared to the number of monthly searches.

I then thought of something grander: what if I could take these ratios, normalize them (force them to fall between 0 and 1) and come up with a weighting system to give me an SEO score for a particular keyword? This would, quantitatively, tell me which keywords are worth using.

My question for the community, in addition to what you all think of this approach, is what weight would you give to the allintitle, allinurl, allinanchor and allintext ratios? Would you weigh them equally? Would certain ratios be weighed more than others?

I appreciate any input. Being technical, I can't help but try and put numbers to situations. I'm curious if this approach is valid, or if I'm simply wasting my time.

[Michael Martinez]

I often advise people that a quick way to assess the competitiveness of a query is to look at the first couple of pages for the query and count up:

• Number of listings whose titles match the query
• Number of listings whose text snippets match the query (as opposed to having words spread throughout the snippet that match parts of the query)
• Number of listings whose URLs match the query
• Number of advertisers for the query

A crude measure for how likely a page is to be optimized for a query is (shown MOST LIKELY to LEAST LIKELY):

• All three elements of a listing (Title, Snippet, URL) contain the query and the listing includes sitelinks
• All three elements contain the query and a second listing from the same site appears below it
• All three elements contain the query
• Two of the three elements contain the query
• Only one of the elements contain the query
• None of the elements contains an exact match for the query

So you can score the sites for competitiveness by this or some other measure and then use that scoring in conjunction with or alternatively to the scoring you might devise for the SERP (Search Engine Results Page) itself.

This is just a method for quick analysis. It's not something to base an SEO practice on but it does provide some guidance on what you may be up against.