The Fundamentals of Merit-based™ Search Engine Optimization Measurement

Web Analytics Association Logo
University of British Columbia logo

Antezeta recently had the opportunity to collaborate with the University of British Columbia and the Web Analytics Association in the creation of a new course on Web Analytics. The full document is available for free to our clients.

This document considers the application of web analytics techniques to the field of merit-based search engine optimization (SEO). We'll look at which dimensions can be measured and some of the techniques and tools available. With an understanding of what can be measured a performance baseline (the current situation) can be recorded. Each measurement dimension can be charted on a regular basis, allowing for objective evaluation of changes.

Contents

[Contents ↑]

Introduction

Search engine optimization (SEO) is often described as part art, part science. Successful SEO requires interdisciplinary skills — skills ranging from the linguistic and psychological challenges inherent in keyword selection to a technical understanding of web page construction. But as most Internet traffic is driven by search engines, few successful web sites can afford to ignore or botch SEO.

Before we discuss measurement and monitoring of SEO, let us establish the context with a quick foundation review.

SEO in a nutshell

Major search engines use complex algorithms, often with over 100 parameters, to determine what results will most likely satisfy a visitor's search request. While dominated by just a few major players, the market is very dynamic – search engines are continuously refining their unpublished algorithms. There is one constant: search engines want to deliver the most relevant, compelling content to their visitors. Merit-based Search Engine Optimization focuses on ensuring a site has:

By Merit-based Search Engine Optimization, we are explicitly focusing on activity which enhances a site for human visitors and by extension, search engines. While there are many not-so-scrupulous techniques which will gain high visibility in search engines, these techniques are counter productive:

[Contents ↑]

Integrating Measures in the SEO Continuous Improvement Process

Organizations with a successful SEO culture implement changes, verify the results and tweak based on the resulting feedback. A new cycle of verifying and tweaking begins. This improvement process is ongoing:

Thus, proper SEO is a continuous, iterative process requiring dedication, patience and perseverance.

What follows is an introduction to some of the measures organizations of all types and sizes can use to validate their SEO efforts. The examples serve as guidance, not prescription. The exact selection of which metrics to adopt and capture, the tools to use and the capture frequencies are best decided by an organization based on its website goals, staffing levels and tool capabilities.

Organizations with short-term needs should consider Pay per Click (PPC).

[Contents ↑]

Metric 1: Search Engine Referrers

When an Internet visitor navigates from page to page, the URL (domain and page, i.e. http://www.theirsite.com/page_which_sent_me.html) they were previously viewing is usually available for tracking by a web analytics system. In the case of a visitor clicking on a link in a search engine, the search terms and the page they requested are available as well.

Once a visitor starts navigating within the site, the referrer contains the site's own domain and the name of the page just viewed within the site.

The most common reason for referrers from other sites is that they spontaneously linked to a site's interesting content. There may also be sites which have agreed to a reciprocal link exchange. Other referrers might be due to participation in paid inclusion (PPC) programs.

In some cases the referrer is empty. This can indicate a visitor entered the URL directly in their browser's address box, they selected it from a bookmark, or they are using software which blocks this information from being sent when requesting a page.

Metric 1. Percent Referrals from Search Engines: A measure of overall search engine optimization effectiveness is to monitor the ratio of referrals from search engines relative to all external referrals.

([Search Engine Page Referrals] / [Total Page Referrals]) * 100

Using the example below: ( 82 / 2781 ) * 100 = 2.9%

Table 1: Page referrals for a given reporting period.
Item Page Count % Referrals
None (Direct entry, bookmark, not provided) 746 26.8%
Links in (Excluding search engines) 1953 70.2%
Search Engines 82 2.9%
Google 68 2.4%
Yahoo 7 0.3%
MSN 4 0.1%
Tiscali 2 0.1%
AltaVista 1 0.0%
Referrals (Total) 2781

All standard web site traffic analysis tools are able to provide data on page requests coming from Search Engine referrals. Successful SEO should drive this number up within a few months after beginning to implement SEO improvements.

Metric 1. Variation

An alternative formulation, for tools which support it, is based on based on visits rather than pages:

([Search Engine Referral Visits] / [Total Site Visits]) * 100

In both cases there is some under-counting of search engine traffic. Returning visitors who originally found the site via a search engine may directly enter the site address (or choose a bookmark) when they return. In such a case, they won't appear as being ‘inspired' by search engine listings, even though this was originally the case.

Web analytics system configuration tip: Insure self-referrals are excluded from your referrer reports. Self-referrals are intra-site traffic.

[Contents ↑]

The use of the term Merit-based™ in conjunction with Search Engine Optimization is a Trademark of Antezeta.

Home · Contact Us · Site Map & Search · Keyboard shortcuts · Top ↑