Mining Web Analytics to Measure SEO Success – Even Without Conversion Data

by Matt Williams - Prominent Placement

There's a continual recognition gap among many marketing professionals when it comes to their websites and understanding the accountability and value search marketing provides. Without a doubt, marketers tend to have a solid "forest view" of search. It's the "tree-level view" they often have difficulty seeing.

What happens then to the marketer charged with proving the success of search engine optimization when a company has been focused on its tree farm and not concerned with actually counting the lumber produced by that farm?

In a perfect scenario, online success is measured by conversions or the number of desired actions taken by visitors to a site – such as a sale, a contact us form, a newsletter sign-up, a white paper download, etc. However, many web analytic programs are not capable of measuring conversions. And, if the company's back-end system is indeed capturing conversion data, it often is impossible to marry this data to search activity.

Perhaps you are saddled with a basic analytics package, one that can measure a lot of site statistics, but usually not conversions. Zero conversion data, now what? How do you demonstrate the value of SEO? You can always produce the standard ranking reports with the growth in rankings and the number of visitors referred by search engines. That's fine – visibility for targeted keywords and increases in search-generated traffic are always good to show, but you're still lacking data showing a visitor did something you wanted them to do.

So when lacking conversion data, you need to produce evidence demonstrating a basic premise of SEO – an ability to drive quality traffic to a website. The majority of basic analytic programs provide much of what you need, and coupled with some basic math can easily produce information that authenticates increases in traffic quality.

To document the key performance indicators (KPIs) for visitation quality, you'll need to track:

Average Time on Site: The longer visitors are on a site, the more engaged they are with the content. This KPI is usually a standard measurement within most analytics programs and should trend upward before stabilizing when compared to levels before site optimization.

Bounce Rate: This is the percentage of visitors who arrive at an entry page, then leave without getting any deeper into a site. These are "one page only" visitors and this is a measure that generally indicates a visitor did not find the content relevant. Some analytic programs specifically report this stat. For those that don't, you can often find a report within the program called "Number of Page Views per Visitor." Since these visitors had only one page view, you can calculate the bounce rate by dividing the number of these one page only visitors by the total number of visitors (e.g., a site with 1,000 visitors during a month that had 400 one page only visitors produces a 40% bounce rate). Compared to pre-optimization, this KPI should trend downward before stabilizing.

Visitors with 2+ Page Views: These are the visitors who arrive at an entry page and progress to at least one additional page on the site – this is the opposite view of Bounce Rate. These visitors were interested in your content enough to explore more than one page, indicating qualified traffic. Using the example above, 600 visitors had two or more page views and represented 60% of total traffic. Compared to pre-optimization, this KPI should trend upward before stabilizing.

Average Page Views per Visitors with 2+ Page Views: More telling than an overall Average Page Views number, this measurement shows average page views among those who actually delved into the site. To calculate this figure, subtract the number of one page only views from the total page views (e.g., a site with 4,000 total page views that also had 400 one page only views would end up with 3,600 views generated from visitor sessions with 2+ pages viewed). Once you have this number, divide it by the number of visitors with 2+ page views. Using our previous examples, we'd divide 3,600 by 600 (2+ total page views divided by 600 visitors with two or more page views) to arrive at a 6.0 average. Compared to pre-optimization, this KPI should trend upward before stabilizing.

Entry Page: The majority of any site's traffic will always come through the home page. However, by dispersing your targeted keywords throughout your site, an optimized site should land visitors deeper and closer to the content visitors are seeking. Tracking this KPI should focus on following the percentage of home page entries which should trend downward before stabilizing when compared to pre-optimization levels, demonstrating the increased visibility of the site for multiple terms within the search engine results pages.

By tracking these "visitor quality" KPIs and comparing pre- and post-optimization results, you should easily be able to prove the value of SEO to your senior management (or client) who hasn't quite had that "a-ha" conversion epiphany. Who knows, they might just catch conversion fever and fork over for a comprehensive web analytics program!

Discuss this article in the Small Business Ideas forum.

Matt Williams is a Managing Partner for Prominent Placement, a search marketing firm in Atlanta. Matt oversees Data Analysis & Reporting, New Services, Operations, and Marketing for the firm.