Performance Implications of Internet Marketing Campaigns on Web Servers: PART I

Internet marketing is the approach of applying advertising techniques to generate traffic to websites with the goal of consuming some service, product, information, etc. Generally speaking, common Internet marketing techniques include (but are not limited to):

• Search Engine Optimization (SEO)
• Pay-Per-Click (PPC) advertising
• Display advertising
• E-mail marketing
• Affiliate marketing
• Social media optimization (SMO)

In recent years, the business community has developed techniques that can drive traffic to web servers; however, little work has been done to develop models that can evaluate the performance of web servers resulting from increased traffic generated by Internet marketing campaigns.

From a capacity planning perspective, this is an important consideration given that Internet marketing is expected to grow at rates of at least 20% per year. For instance, some Internet marketing approaches employ PPC, which when implemented can generate traffic to a website in a matter of minutes. Others may take longer like SEO.

In these series of posts, we present traffic models that are able to estimate the traffic due to Internet marketing campaigns:

• A surfer click-through model
• A content building model
• A traditional capacity planning model that uses historical data to predict future traffi

The traffic models can then used as input to an M/M/1 queue where web server performance measures can be ascertained. Our empirical results suggest that Internet marketing can have a significant impact on web server performance – much more than expected. Thus, it’s important for system administrators, web hosting services, etc., to understand the relationship between Internet marketing and web server performance since user-perceived performance is often a major factor determining whether e-commerce transactions are completed.

In the next post, we’ll discuss how SEO traffic models can be developed and utilized to assess the performance of web servers.