In Q1 of 2010 we were approached by Internet Retailer Magazine and asked to develop a methodology for evaluating the SEO effectiveness of the Internet Retailer 500. In this year’s annual Internet Retailer Top 500 Guide Conductor’s ‘SEO Score,’ developed using the methodology described below, appears on each retailer’s profile page providing readers with a standardized measure of the retailers SEO expertise.
In the Internet Retailer Top 500 Guide and Search Marketing Guide, you’ll find Conductor’s ‘SEO Score.’ It was developed using the methodology described below, and appears on each retailer’s profile page providing readers with a standardized measure of the retailer’s SEO expertise.
Here at Conductor, we’ve spent a lot of time thinking about SEO as we help our clients with solutions to achieve visibility in natural search. Yet as we set about considering how best to gauge a firm’s SEO effectiveness, we concluded in an ideal world, a firm’s SEO efficacy would be measured by analyzing the analytics for the keywords the firm is attempting to optimize.
Given the understandable reluctance for organizations to share this info externally, we sought to develop an externally-verifiable, metrics based approach to measuring their relative SEO effectiveness and commitment. How, we wondered, could we obtain a holistic view of the SEO effectiveness of an organization from the outside in?
We began by considering the elements an organization must be good at to be considered proficient at SEO. We determined there to be three key areas where a firm must excel, and developed a methodology for measuring each while external to the organization:
The human resources a firm dedicates to SEO (relative to their overall resources) is an indicator of their level of commitment to search engine optimization—without the resources to manage and optimize keywords the organization will be unable to administer an ongoing SEO campaign of any scale. For each retailer, we gathered the employees on Linked-In who’s job function includes search engine optimization. The number of SEO employees was then divided by the total number of the firm’s employees listed on Linked-In for a percentage of employees who feel SEO is a large enough portion of their job description to include it in their profile. As an acknowledgment of a firm’s commitment to SEO we awarded a 10% bonus for every Director level SEO employee and 15% for every VP level or higher. A ‘Resources Score’ was assigned to each retailer on a sliding scale based on their final percentage.
The ability to measure the impact SEO efforts are having on site traffic is central to professional management of an SEO campaign of any scale. Using external tools we determined the analytics package the retailers had installed on their website. Retailers with no analytics received a ‘0’, those with Google Analytics received a ‘medium’ score and those with a commercial analytics package received a ‘high’ score. While some small percentage of retailers may be using post-visit log analysis as opposed to externally visible analytics packages, this should be considered a low number.
As a final measure of the retailer’s SEO effectiveness we devised a method to gauge how effective they are in ensuring the visibility of keywords that matter to them. Absent access to the list of target keywords each retailer is attempting to surface to the top of search results, we measured their intent by assuming they voted with their wallets in prioritizing their paid keywords. We gathered each retailer’s 200 most expensive paid keywords from Internet Research Firm SpyFu and tracked where on Google’s search engine results page (SERP) each of the 200 keywords appeared. Each keyword/domain combination was given a grade–with the highest grades going to companies who appeared in the first 10 listings, and the score depreciating as the search visibility deteriorated. The average of the scores was taken to arrive at a ‘Visibility Score’ for each retailer.
A final SEO score was calculated by summing the three scores. In the small percentage of instances where one of three data points was unavailable for a retailer (e.g they did not have any employees on Linked In) the average of the other two scores was used, and this is noted on their score. Score classification thresholds (‘excellent’, ‘good’, ‘fair’, ‘poor’) were determined by dividing the top score by four.