Measuring SE ranking results

Measuring SE ranking results is controlling and reporting the changes to any website's positioning status. A good ranking in the search engines provides a steady supply of interested visitors/customers, while a poor one sinks any idea into oblivion, no matter how good or relevant.

There is a whole industry called SEO, search engine optimizing, dedicated to obtain good web rankings. The first issue that appears when you consider hiring a SEO company is how to measure the web rankings, before and after the company did his job.

It is also valuable to assess the ranking value of different clients and campaigns performed by a certain company, operator or methodology. Thus, measuring the positioning of a web page in the search engines is very useful for site owners, web marketers, programmers specialized in SEO and many other web-related professionals.

In spite of this need for measuring rankings, there is no consensus among the webmaster community on how to do it. SEO Contests are a very specific way of comparing the ability of SEO experts, but it does not measure SEO ability outside the given keyword, timeframe and rules.

This proposal of a unique Ranking index is an improvement over earlier versions of the SEPI (search engine positioning index).

SE do not reveal their ranking algorithm, although apparently there are over 200 factors taken into account, being keyword frequency and linking structure the most influential. The results however are stable enough in time to be analyzed and compared.

The results of different strings of text contained in a web page are extremely variable. The most useful search string is usually the Title metatag, because it has the most chances of being ranked.

The positioning value of a website is the combination of its individual keyword rankings, pointing to any page within the site. So, a valuable index needs to be additive, combining all frequent queries directed to pages within the site.

The ranking results depend on how the search is performed. Proven factors influencing ranking of a chosen keyword in searches are:

   * Length of the selected keyword phrase
   * Presence of quotes
   * Language
   * Location of the query (as detected by the connection IP)
   * Data of the eventual user configured in Google or Yahoo (currently, the SEs try to provide results relevant to the user, based in its location and language preferences, and maybe other factors)
   * Time of the query (sometimes the results vary slightly in time)

These factors have prevented so far the establishment of an accepted index for SE results. However, some simple math can overcome this problem.

In order to define a reproducible index, we need to establish a few rules:

   * Google.com is the default SE
   * The SE is configured without a user.
   * Language is English.
   * Searches provide 100 results.
   * SERP (search engine result pages) count all the indented pages within a site.
   * When the keywords have a local component, like “Manhattan Pizza”, the searcher IP needs to be logically related to it. Local keyword phrases should be avoided when comparing rankings of different location-related keywords.
   * The searched keyword phrase for a given page must be contained in the Title.

The SEPI can be calculated with or without quotes. Without quotes there are more total results and ranking is more difficult.

For now, we will use SEPI 1.0:

SEPI 1.0 = Total SERP Results under certain keyword / (position of the keyword 2 x Keyword String Word Length)

The SEPI can be calculated with or without quotes. Without quotes there are more total results and ranking is more difficult.

The number of SERP Results is the easiest value to obtain after a search, and a very good indicator of the Keyword Difficulty.

The position of the keyword is squared in the formula because it is very important to be near the top. The 30 first results in any search are usually taken into account, while anything below the 3rd or 4th page has almost zero chances of being visited. Since the most common searches bring millions of results, there is an exponential decrease of the positioning value going from top to bottom of the SERP.

The Keyword String Word Length is included in the formula, because shorter strings are more frequent and thus more valuable.

Keyword Difficulty can be measured with more detail, although for a general index it is not really necessary. There are tools that measure keyword difficulty ^[1], considering the number of existing pages, the PageRank of the top-ranked pages and maybe other parameters.

The SEPI is intended to be applied to keywords of any difficulty. There should be a correlation between SEPI for a standard page and Keyword Difficulty: the more difficult the keyword, the poorer the ranking results. Of course, there is not such a thing as a Standard Page, because minor changes in keyword frequency, text length, linking and many other factors affect the rankings. We have designed pages with similar parameters and different keyword difficulty, and apparently there is a negative correlation.

Another verification of the value of SEPI would be the prediction of SE traffic coming to a ranked website. The traffic would be a function of the Keyword Popularity (number of monthly searches as provided by Google Keyword Tool) and the total site SEPI (addition of important SEPI values for common keywords).

A possible experiment will demonstrate the quality of this and other candidate formulas: 1) select domains for a niche subject. 2) discard those that have incoming links as detected by Yahoo. 3) obtain the search frequency for the relevant keywords. 4) estimate the traffic from third sources, like Alexa. 5) measure the quality of SEPI to predict traffic.

This issue is open to discussion. Suggestions for improvement will be incorporated in future versions of SEPI. A current free SEPI calculator is provided at Search Engine Positioning Index [1].

References

1. ^ SEO Tools "Website Optimizing Tools"- Website

External links

• "Search Engine Ranking Factors"