Foreword

One of the most interesting aspects of web site measurement is that it overlaps with so many areas of a company’s business. A company’s web presence covers the whole of its interaction with the public. It is marketing and sales and customer relations and press office and recruitment. It has to inform the public, create new customers, and support existing customers.

Web analytics is the study of whether the web site is meeting its diverse goals. Just as important, it is the presentation of the results to the various divisions of the company in a comprehensible format. It is now an industry worth hundreds of millions of dollars annually.

It wasn’t always that way. I started writing Analog in 1995 when I was a student at the University of Cambridge Statistical Laboratory. Our department had just set up a web site, and we were keen to know how many people had been visiting it, but we found that none of the existing three programs worked well. So I decided to write my own little program, not guessing how it would take off.

Back then, the terms “web site measurement” and “web analytics” hadn’t yet been coined: it was usually called something like “web logfile analysis.” And the programs had a different emphasis than today’s commercial programs: they focused on more technical statistics, such as which pages had been viewed most often and how many bytes had been transferred, rather than on visitor behavior.

Several things happened to broaden the scope of web measurement. The most obvious is the growth in commercial activity on the Web in the late 1990s. As the Web became a major part of a company’s business and a major expenditure, it became necessary to justify that expenditure. This perhaps became even truer during the weaker economy of the last few years, as all expenditure had to be examined, and the Web came to be regarded as a marketing channel like any other.

Another important development was the growth in pay-per-click advertising in the last four or five years. When Overture and Google introduced the ability to place text ads on search engine results pages, it brought web advertising within the reach of many more companies. Compared to traditional banner ads, these new ads were better targeted, and charged only for actual clicks. They were also self-service, which made them cheap to set up and easy to change, and allowed companies to experiment with many different ads. Just as the Web gave everyone the ability to become a publisher, pay-per-click ads gave everyone the ability to become an advertiser.

One technical development also deserves a mention. JavaScript was invented in 1995 as a way to embed small programs within web pages. Its relevance to web measurement is that a piece of JavaScript code can alert a dedicated data-collection server when the page is displayed. This allowed vendors to offer measurement as an outsourced service, instead of as a software purchase. There is an ongoing debate as to whether the JavaScript method or the traditional logfile method is superior—each has advantages and disadvantages, as this book will discuss—but it is certain that JavaScript made web measurement available to many companies with less technical expertise, and to those whose web sites were hosted on third-party web servers.

In this environment, where companies were spending large amounts of money on their web sites and needing to examine the expenditure, web measurement vendors began to focus less on technical statistics, such as browser types and bytes transferred, and more on commercially relevant statistics, such as conversion rate and return on investment. The web measurement field gradually changed emphasis from “server analytics” to “visitor analytics.”

In addition, vendors invented new ways to present data to wider audiences. Web statistics are useful to a business only if they can be understood by the people who have the authority to change the web site. If the statistics cannot be understood by people who can act upon them, they are merely an expensive curiosity. It was this that attracted me to making a career, not just a hobby, out of web analytics. I had refused previous job offers, but I joined ClickTracks because they wanted me to develop innovative ways to present the data—ways that were intuitive but still with an underlying mathematical rigor.

In this book, you’ll learn techniques to use today’s web measurement programs most effectively, written by many if not all of the leading experts in the field. To be successful, every modern business needs to understand the behavior of customers and potential customers on their web site; and if they do, they can see substantial reductions in costs and increases in revenue. It is the aim of the authors to give you new insights into the visitors to your web site: insights that will directly improve your business.

—Dr. Stephen Turner, Cambridge, England

Get Web Site Measurement Hacks now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.