Strategy for Optimizing Domain Name Monetization Revenue

Alex Tajirian
April 5, 2006

Introduction

Designing and implementing new techniques to increase a domain name’s traffic monetization revenue is a never-ending endeavor. However, a strategic approach needs to be implemented to properly evaluate results. The paper outlines current methodologies to analyze the impact of optimization, puts forth a new factor-model approach, and points out some of their strengths and limitations.

Approaches

There are two distinct methodologies to evaluate an action: event studies and modeling the revenue generating process. The former analyzes the effect of an event or action on revenue. An event may be a new landing-page design or modification to its word content.[1] The latter approach, on the other hand, is a model that is driven by systematic factors that generate monetization revenue.

One of the main drawbacks of event studies is that it is not easy to isolate the impact of events. For example, changing the design of a landing page will most likely not produce systematic revenue changes across domain names. Thus, the causes of differences in revenue response to the same action remain unanswered. Moreover, the investigator needs to let time pass before he or she can measure the impact on revenue of the redesigned landing-page. However, during the waiting period, the world does not stand still and other factors enter to influence revenue. Furthermore, anecdotal evidence suggests a tremendous focus by practitioners on search engine optimization (SEO), which may, in reality, have a very small contributing factor to total revenue.

Unfortunately, however, there are no theoretical models to pre-identify the revenue generating factors. Nevertheless, there are two promising approaches: one is based on the selection of intuitive and parsimonious factors, and the other on unobservable factors. (Yes, Unobservable!) The approaches are outlined below:

Approach 1:  Postulate a two-factor model with the factors being branding and traffic.[2] Such a model has tremendous benefits; it makes it possible to understand the contribution of each of the components on value and thus, provides a basis to assess the best use of a domain name, domain selection, the viability of branding protection with TLDs, and distinguish between good and bad traffic.[3]

Approach 2:  Instead of explicitly specifying factors, statistical techniques, such as principal components, can be used to determine the number of systematic factors explaining variations in revenue among various domain names and then the researcher can correlate the systematic factors with measurable factors.[4] Thus, one would then be able to examine the effect on revenue of changing one factor at a time, while controlling all others.<<