I am implementing the Keyword Effectiveness Index (KEI) into the Maverick Keyword Researcher. However, the original formula does not work with Google keyword statistics unless I make some tweaks to it.

First up, let’s take a look at the official explanation of the KEI by its creator, Sumantra Roy:

The KEI is a measure of how effective a keyword is for your web site. The derivation of the formula for KEI is based on three axioms:

1. The KEI for a keyword should increase if its popularity increases. Popularity is defined as the number present in the “Count” column of WordTracker. This axiom is self-explanatory.
2. The KEI for a keyword should decrease if it becomes more competitive. Competitiveness is defined as the number of sites which AltaVista displays when you search for that keyword using exact match search (i.e. you should use quotes around the keyword). This axiom is also self-explanatory.
3. If a keyword becomes more popular and more competitive at the same time such that the ratio between its popularity and competitiveness remains the same, its KEI should increase. The rationale behind this axiom requires a more detailed explanation. The best way to do this is to take an example:

Suppose the popularity of a keyword is 4 and AltaVista displays 100 sites for that keyword. Then the ratio between popularity and competitiveness for that keyword is 4/100 = 0.04.

Suppose that both the popularity and the competitiveness of the keyword increases. Assume that the popularity increases to 40 and AltaVista now displays 1000 sites for that keyword. Then the ratio between popularity and competitiveness for that keyword is 40/1000 = 0.04.

Hence, the keyword has the same ratio between popularity and competitiveness as before. However, as is obvious, the keyword would be far more attractive in the second case. If the popularity is only 4, there’s hardly any point in spending time trying to optimize your site for it even though you have a bigger chance of ending up in the top 30 since there are only 100 sites which are competing for a top 30 position. Each hit is no doubt important, but from a cost-benefit angle, the keyword is hardly a good choice. However, when the popularity increases to 40, the keyword becomes more attractive even though its competitiveness increases. Although it is now that much more difficult to get a top 30 ranking, spending time in trying to do so is worthwhile from the cost benefit viewpoint.

A good KEI must satisfy all the 3 axioms. Let P denote the popularity of the keyword and C the competitiveness.

The formula that I have chosen is KEI = P2/C*1000, i.e. KEI is the square of the popularity of the keyword multiplied by 1000 and divided by its competitiveness. This formula satisfies all the 3 axioms:

1. If P increases, P2 increases and hence KEI increases. Hence, Axiom 1 is satisfied.
2. If C increases, KEI decreases and hence, Axiom 2 is satisfied.
3. If P and C both increase such that P/C is the same as before, KEI increases since KEI can be written as

KEI = P2/C*1000 = P/C * P * 1000. Since P/C remains the same, and P increases, KEI must increase. Hence, Axiom 3 is satisfied.

You may be wondering why I am multiplying the figure by 1000. The reason is that since the popularity will be very small compared to the competitiveness, the KEI for most of the keywords will be in decimals. This makes it slightly difficult to use. However, introducing the 1000 makes it easy to compare the KEI between different keywords.

Note that the formula for KEI is not unique. In fact, this is one of the nice things about the KEI. If, instead of using 2, you use any power of P greater than 1, the resultant formula will also satisfy the 3 axioms. For example, P1.5/C*1000 and P3/C*1000 both satisfy the 3 axioms. The exact power of P that you choose depends on how much emphasis you want to give to the popularity of a keyword viz-a-viz its competitiveness. Higher the power of P in the formula, higher will be the emphasis on popularity. If you are very confident about your search engine positioning skills, choose a higher value for the power of P.

If you are not that confident about your search engine positioning skills, choose a lower value for the power of P (but the power should still be more than 1). Thus, the KEI can be adapted to your skill level! Feeling confused as to which power you should choose? Stick to 2. It maintains a nice balance between both popularity and competitiveness.

======================

You can use the concept of KEI or any derivation of it in your own articles as long as you display the following resource box at the end of the article:

The Keyword Effectiveness Index was invented by Sumantra Roy. Sumantra is one of the most respected and recognized search engine positioning specialists on the Internet. For more articles on search engine placement, subscribe to his 1st Search Ranking Newsletter by sending a blank email to 1stSearchRanking-subscribe@listbot.com or by going to http://www.1stSearchRanking.com.

======================

Keyword Effectiveness Index (KEI) with Google Keyword Data

I hope that wasn’t too much math for you to understand. The KEI was created quite a few years ago while Altavista was still some kind of big shot in the search engine industry. Today, Google is the king of the search engines so we need to use Google’s data instead of Altavista and Wordtracker.

I tried using the original KEI formula with Google keyword date and the numbers looked unbelievably good. So good that is must be false.

Let’s replace all the Competion and Popularity jargons with something easier to read and remember. So here’s the KEI adapted to Google’s keyword data:

KEI = Google Search Count² / Google Search Results x 1000

Let’s use the keyword phrase “internet marketing” as an example. I shall use the exact match to search for internet marketing at Google.com. Please note that exact matching on Google.com refers to the use of quotes around the keywords wheareas brackets [] are used for exact matching on Google AdWords. Please refer to the Google Web Search’s search basics and Google AdWords’ keyword matching options for more information on the differences between Google’s matching options on Google.com and Google AdWords.

My search settings are as follows:

• Country: USA
• Language: English
• Google.com exact matching
• Google AdWords exact matching

I get my keyword data from the Google AdWords Keyword Tool and Google.com. I shall use the Global Monthly Search Volume divided by 30 for my Google Search Count and the number of results I see in the top right hand corner of Google.com’s search pages for my Google Search Results. The original formula uses daily search counts from Wordtracker so it is best that I use a daily search count from Google to get the formula right.

So here’s what I’ve got…

KEI = (201000/30)² / 44400000 x 1000
KEI = 1011.04

1000 is a really good KEI index. In fact, it is too good to be true. It should not be that easy to rank for the term “internet marketing”. The problem with this formula is the x1000 part at the end. If you remove that, the formula seems to make more sense. Google is a monster compared to Altavista and I believe there is no reason to multiply the results by 1000.

What if we don’t multiply it by 1000? Here’s what the KEI index will be:

KEI = (201000/30)² / 44400000
KEI = 1.01

I prefer the tweaked formula because the number returned makes more sense to me.

Updated Keyword Effectiveness Index (KEI) Formula for Google

If you are planning on using the KEI for your web site’s SEO research, here’s my updated KEI formula for your quick reference.

KEI = Google Search Count² / Google Search Results

There could be a better way of calculating the KEI. Some people use the “allintitle” search parameter to get different results. If you know a better tweak for Google keyword data, please let me know.

User Accounts, Installation, Password Reminders, Login etc

The first public version of the Maverick Keyword Researcher will not use a MySQL database to store its data. I have developed it to use text files for the time being. It is easier to sort keywords with MySQL than text files. So the first version will have limited sorting features.

The installer is pretty straightforward. You upload all the files to your web server or your local server and launch index.php. The installer will make sure you have certain files and folders with the right permissions. Since we will be using text files in version 1, we need to set the file permissions of a couple of folders and the text files in them. This is the troublesome part but it is easier than creating a database in cPanel.

Maverick Keywords Researcher will be encrypted with ionCube’s PHP Encoder. The industry’s top 2 PHP encoders are Zend and ionCube. Most web hosts have them installed by default. If not, they should be able to install it on your web server for you if you ask them to.

I am thinking of making it more user-friendly in future releases by using ionCube’s Package Foundry. This software will help you to upload and and configure all the folder and file permissions. It may be able to install the ionCube loaders on your web server if you don’t have it installed. I will purchase ionCube’s Package Foundry with the initial sales of this product. I also need to upgrade my ionCube PHP Encoder to a better version to generate license files and stop the licenses of customers who have refunded. I could also give away free trials with limited licenses.

Google Keyword Research

I combined a few Google keyword tools on a single page. I am working on adding Google’s Search-based Keyword Tool too. At the moment, my Google tool grabs related keywords from Google.com home page (Google Suggestions), Google’s related searches, Google Synonyms and Google Insights for Search.

I discovered a way to generate more keywords by searching multiple countries. Most countries will give you duplicate results but if you select the most popular countries (I got this list from the countries listed at the top of Google’s country drop down lists…), you should be able to get a very nice list of keywords. If you’re up for it, you could select all 233 Google-supported countries.

I am still working on getting keyword suggestions from Google AdWords. The Google AdWords Keyword Tool has to be on a different page because it requires a captcha. It would be irritating for the users to mix it with the other Google tools that don’t require a captcha.

KwBrowse Keyword Research

This KwBrowse tool took me a long time to complete. I could not find a way to automate it completely because it returns a lot of unrelated keywords. I created a select box on the right to filter out the unwanted keywords that are returned from the searches.

It isn’t easy or fast to get a complete keyword list from KwBrowse but it is worth the effort because you will get a huge list of related keywords.

Wordtracker Keyword Research

Wordtracker has a free keyword tool that generates the top 100 related keywords of the keyword you enter in their form. Sometimes, this keyword tool requires a captcha before you are allowed to search. I have programmed the Maverick Keyword Researcher to deal with that issue. The problem with a lot of free keyword tools online is that they don’t give you an easy way to copy and paste the keywords. With my tool, they are returned nicely in a text file.

Others…

I created a list of categories that can be easily copied and pasted for mySimon and Shopping.com. You could grab these keywords and find potential niches for your online business. This list has over 500 keywords and you should be able to find a few profitable niches in there.

I also created keyword suggestions from Alibaba, Ask and Yahoo. Some of the keyword suggestions cannot be copied and pasted in the browser. For example, if you enter a keyword at Yahoo.com or Google.com, the web page returns a list of related keywords in your browser. The only way to get that list manually is to type the keywords you see on the screen to some application on your computer. That’s just stupid and a waste of time. I figured out a way to get these keyword suggestions quickly and easily.

Alibaba, Ask, Yahoo and Google all have a suggestion list when you type in a keyword at their home page. Ask.com has additional keyword suggestion on their search pages and I have added those into the Maverick Keyword Researcher too.

I am thinking of adding keyword suggestions from eBay and DHgate.com. Another source would be Keyword Discovery. For version 1.0, all the keyword sources would be free. I might include paid keyword sources from Wordtracker, Keyword Discovery and Wordze in the future. Maybe I could make it a monthly subscription thing at a lower rate than the amount you pay for these keyword tools every month.