Understanding keyword efficiency index (KEI) - My Thoughts

by Fary
9 replies
  • SEO
  • |
Hi warriors,
If you are wondering whether to rely on KEI for measuring the success of any keyword or not, then this topic is for you. I am sharing some advice from experts along with my own thoughts so we can discuss them.

First of all, KEI is a measure to say if a keyword is good to rank for or not. It depends on the following factors:
- Relevance (R): This is an estimate of how your keyword is really related to your niche. You should give the keyword a number from 1 to 3 (excellent, good, bad).
- Search Volume (SV): how many search queries per month for this keyword.
- Competition (C): how many websites are competing with you to rank for the keyword.

In short: R is a factor you decide yourself, SV is a factor given by Keyword Planner, C is a factor known by simply searching for the keyword to see how many websites appear.

KEI is calculated as follows:
KEI = ((4 - R) / 3) * ((SV * SV) / C)

Example:
R = 1 _ SV = 10,000 _ C = 2,000,000
KEI = ((4-1) / 3)) * ((10,000 * 10,000) * 2,000,000) = 50

Those of you who know how to apply this formula to excel sheets will find it easy to calculate.

Now, as we already know how to apply the formula and taking in account the fact that KEI depends on Relevancy so it is some sort of 'smart tool'. It does not only depends on raw data like visits and competition but also cares about the 'quality' of these visits.

I personally found this helpful for key phrases not for single keywords. A single keyword may give a very high KEI but in reality it is merely impossible to rank in it. In 2 and up to 4 key phrases I found it helpful.

Now, would you depend on KEI for evaluating your keyword plan? If not, what is a better choice?
#efficiency #index #kei #keyword #thoughts #understanding
  • Profile picture of the author webblo
    I will keep in my mind. Thx
    {{ DiscussionBoard.errors[9656867].message }}
    • Profile picture of the author Fary
      So you find it really helpful, webblo?
      {{ DiscussionBoard.errors[9656873].message }}
  • Profile picture of the author MikeFriedman
    You are a complete moron if you use any metric that takes into account how many webpages show up in the search engine index for a search term. That has NOTHING to do with the level of competition.
    Signature

    For SEO news, discussions, tactics, and more.
    {{ DiscussionBoard.errors[9656870].message }}
    • Profile picture of the author Fary
      Originally Posted by MikeFriedman View Post

      You are a complete moron if you use any metric that takes into account how many webpages show up in the search engine index for a search term. That has NOTHING to do with the level of competition.
      Thanks for your advice, Mike. Is that true even if you use double quotes "" to evaluate websites that use the EXACT keyword as you?
      {{ DiscussionBoard.errors[9656876].message }}
      • Profile picture of the author MikeFriedman
        Originally Posted by Fary View Post

        Thanks for your advice, Mike. Is that true even if you use double quotes "" to evaluate websites that use the EXACT keyword as you?
        Yes. It still tells you nothing about the competition.

        It doesn't matter if there are 50 webpages or 50 million webpages. If you cannot outrank any of the sites in the top 3, who cares what #4,973,210 is doing?

        You have 3 competitors. That's it.

        For really high volume keywords, maybe 5. Those are the competition.
        Signature

        For SEO news, discussions, tactics, and more.
        {{ DiscussionBoard.errors[9656879].message }}
        • Profile picture of the author Fary
          Originally Posted by MikeFriedman View Post

          Yes. It still tells you nothing about the competition.

          It doesn't matter if there are 50 webpages or 50 million webpages. If you cannot outrank any of the sites in the top 3, who cares what #4,973,210 is doing?

          You have 3 competitors. That's it.

          For really high volume keywords, maybe 5. Those are the competition.
          So what is a good alternative to quickly measure competition and sort a big number of keywords accordingly?
          {{ DiscussionBoard.errors[9656884].message }}
          • Profile picture of the author MikeFriedman
            Originally Posted by Fary View Post

            So what is a good alternative to quickly measure competition and sort a big number of keywords accordingly?
            There is no quick way to do it efficiently.
            Signature

            For SEO news, discussions, tactics, and more.
            {{ DiscussionBoard.errors[9656920].message }}
  • Profile picture of the author danparks
    Originally Posted by Fary View Post

    - Competition (C): how many websites are competing with you to rank for the keyword.
    As pointed out, the number of search results displayed by Google is meaningless for checking competition. I randomly picked a ridiculous keyword to check:

    dogs with bad breath in detroit

    There are over 3 million results. Who cares?

    In the end, all that matters is who exists at the top of Page 1 of search results. You want to study those few pages and see if it's realistic for you to beat them with whatever page of yours you're attempting to rank.
    {{ DiscussionBoard.errors[9657442].message }}
  • Profile picture of the author Work1099
    This algorithm is suffering in one key area. It's calculating competition based on the number of competing websites. Not all competing websites are real competition; most of the time, there are a few people who know what they are doing, and a vast mob who have no clue. The only real competition is the people who know what they are doing ... the others can be out-done easily.

    Also, this algorithm would be far more accurate for comparison if it not only focused on only real competition instead of total websites ... but also if it took into account the difficulty of beating those other websites. Even if the difficulty is only estimated on a scale of 1 to 10 (and what is chosen is based on detailed analysis, not a first impression), it would still be a vast improvement in terms of comparing keywords.
    Signature


    Work1099 Simplifies Self-Employment.

    Get 600+ Proven Business Models and Money-Making Ideas at: http://work1099.com
    {{ DiscussionBoard.errors[9659168].message }}

Trending Topics