A keyword-independent measure of search engine positioning

So far, it is not possible to compare multi-keyword positioning (ranking) results in google or yahoo. Positions are not additive, because every keyword has a different difficulty. Obviously, it is better to be 1st of 100 than 1st of 10.

However, is it better 1st of 100 or 2nd of 1000? How about any other non-obvious pair of rankings?

Moreover, things get complicated when there are more variables: in addition to position (rank) and total number of SERP (search engine result pages), the search engine and the specific keyword.

The ideal measure of search engine positioning should have these properties:

– easy to calculate

– additive

– representative of the difficutly and merit of the positioning

– correlate with commercial results for the ranked site

– reflect the habits of the search public

It is known that most people search Google or Yahoo for the first 5-6 results. However, to cover the whole population, we need an idea of the math explaining this behavior, like an asymmetic Poisson distribution curve.

When several SEO companies want to compete, they create a contest with a single keyword or keyphrase: ‘mangeur de cicogne’, ultramarine nigritude and other nonsensical, unique, rarely sought phrases. This method is cumbersome, takes time, and otherwise useless. And it is completely separated from real life situations.

It would be much better to have a method to compare existing positioning results.

Thus, I proposed the SEPI, Search Engine Positioning Index, as:

 SEPI = Total SERP x Keyword Difficulty ^2 / Position ^2

Keyword Difficulty can be estimated by the PageRank of the tenth page for a Google search on the keyword. 

You can use this online calculator

This index is useful for:

– comparing results of SEO software or SEO campaigns

– comparing SEO companies.

– charging customers by SEO results

I need consensus from the SEO webmaster community, or suggestions to improve this index.

Our AlgoCracker Software

We are preparing different experiments with the current version of AlgoCracker. We first want to establish the value of different TLDs on Google ranking. If non-standard TLDs are not discriminated against, domains with keywords should be very valuable for ranking purposes.

The problem with TLDs from different countries is that we cannot establish the value of page quality. The .com pages tend to be of better quality, because they are more demanded and thus more expensive.

Which keyword should be analyzed? It should be a keyword of neutral value to the TLD. 

To solve that, we will use the names of the capitals of countries.

More later.

Viral Business Proposal: Complaint boards as means to enforce website service quality

The original idea was a “CompanyBoard” business, in which I would run a Complaint Book for my clients. This Book would have been a forum system hosted in my server, where customers could complain or endorse the company.

The owner of the Complaint Book would supply a link in to the forum in my domain. My hosting the forum would prevent the owner from doctoring the results.

However, I found some complexities related to judge posting quality. Company enemies ccould be negative in public or paid supporters could be over positive.

I still think the business is feasible, but I would not use a forum, but a Digg-like News Aggregator. I am running one named Business-ideas.com.ar, expecting that the best business ideas will climb to the top. I am also running one for political ideas in Argentina: Ideaspol.com.ar.

The news aggregators have a sophisticated voting system that would allow users to write reports about a company, and the community would vote about quality of the report.

The users develop ‘karma’, a quality related to credibility. Karma is automatically managed by a hidden algorithm in the system.

The pligg module will be modified to improve the quality of the complaints, preventing fraud or manipulation.

The companies will be offered a free or paid service. The paid service will have many more features.

This idea is ready to be launched. Today, Oct. 22-2007, I need a partner bringing U$D 6000, for 40% of the business.

I am listening…

New blog platform – WordPress instead of Bblog

The aging ‘bblog’ platform is being decommissioned. It is not being supported since 2005, and we need to change. WordPress has more friendly URLs, more security, no use of global variables, more plugins and so on.

We intend to write in here short abstracts of the better postings in the old system, to avoid losing the old addresses. Eventually we will announce a permanent redirect in our .htaccesss file, which is the better form to explain the changes to the search engines, without losing ranking.

What is Algo Cracker?

It is a module in a software that used to be named Keyword Thief. This product started as a spider that extracted keywords from the top-ranked sites in Google.

In a second stage, it incorporated 3 modules:

– Metatag Thief, which does the same as Keyword Thief, but for all metatags

– Algo Cracker, an analyzer of keyword density in different parts of the Google top-ranked websites. It uses ranges (clusters) of pages, to allow averaging and comparing

– Metatag Analyzer, a spider that enters a single site, and checks the contents of all metatags.

This products are now sold together, but can later evolve separately.

See http://keywordthief.com.ar for details and free download.