Design & Development

Web Analytics: Data Collection Methods

Web analytics software helps ecommerce operators understand what their online visitors are up to. Which search engine did the visitors come from? How long did they remain on the site? Which web pages did they exit the site from? And so forth.

Utilizing the data that web analytics packages provide can help an ecommerce operator improve his business. Understanding how the data is collected can help an operator understand web analytics.

“There are two ways in which web analytics packages collect data,” says John Marshall, President of ClickTracks, a web analytics provider. “Each method has its advantages and disadvantages. An ecommerce operator needs to understand each of them to follow how his particular analytics package works.”

The first approach, according to Marshall and other experts, is known as the “log file” method. This process refers to the trackingfiles that are routinely stored on a web host’s server. These files automatically record visitor behavior (such as time on site, pages visited, exit pages and much more). Hosting companies and webmasters use these files to manage storage and bandwidth issues. But, as Marshall points out, the log files can also be parsed and analyzed by software and the data produced by that software could help web site owners improve their businesses.

The second approach is the “JavaScript method”. This method does not require log files at all. Instead, it relies on JavaScript code that is included with each web page. The JavaScript sends visitor activity to a computer that is hosted by the web analytics service provider. The site owner then uses a client viewer or web browser to view the processed analytics for the site.

John Marshall, “The log file method tends to be less accurate than the JavaScript method. But, since the JavaScript method usually involves paying monthly fees to an analytics company, the log file method is typically less expensive. With the log file method, a website operator typically incurs a one-time software expense, but he can then analyze the log files at no additional cost.”

Marshall also points out that log files can exist whether a website operator uses them or not. In that respect, the operator does not have to change his site and add extra code to his site. This is different than the JavaScript method, which requires additional code to be added to each page of a site and may require other programming changes to it. Also, Marshall adds, the log file method can track search engine spidering, where the JavaScript method typically cannot. Many ecommerce firms, however, rely on the expertise of analytics companies. Those companies almost universally rely on the JavaScript method. Tim Kauffold is Director of Business Development for Oneupweb, a search marketing company that also consults on analytics issues with ecommerce firms.

“We prefer the JavaScript method for our clients because the data is real-time, and we can instantly adjust a client’s site or a client’s keywords if we need to,” says Kauffold. “The data from log files sometimes takes up to a day or two to access, which is frequently too long to wait.”

Kauffold agrees, however, that the log file method has its place and that some of the logfile data, such as exit pages and conversion percentages, is very reliable.

Both Marshall and Kauffold agree that the JavaScript method raises privacy issues. That’s because it relies on the placing of cookies on an (oftentimes) unsuspecting visitor’s computer, and then allows an independent company to store and review that data. Although most web analytics vendors remain indifferent to their customers’ data, a few are beginning to use customers’ visitor data to issue press statements about conversion statistics and do not rule out its use for advertising purposes.

Complicating these privacy concerns is the use of “first party” cookies and “third party” cookies. First party cookies are set directly by the website itself. For example, if, say, directly provides a cookie for, say, Bob’s web browser when Bob visits, that cookie is a “first party” cookie. If, however, hires an independent company who itself provides a cookie for Bob’s browser when Bob visits, that cookie is called a “third party” cookie. In that instance, the independent company presumably saves the data collected by the cookie and, presumably, has less control over that data. Third-party cookies, many experts conclude, protect visitors’ privacy less than first party cookies.

Enter Google Analytics to further complicate privacy issues. Google Analytics (formerly, Urchin) uses the JavaScript method. Google Analytics is free for most companies and Google itself necessarily stores data collected by Google Analytics so that website operators who use Google Analytics can access the data and make decisions from it. But, many experts question whether Google’s search engine also benefits from the data collected by Google Analytics, and whether this breaks most web site privacy policies. For example, experts question whether Google Analytics can provide conversion data to Google Search from advertising keywords that were purchased by a website owner., or whether Google can tailor individual ad search results based on data from Google Analytics.

In any case, Marshall suggests that ecommerce firms at least understand some of the privacy questions.

“Website operators should be aware of these privacy issues,” says Marshall. “Most ecommerce privacy policies do properly address the fact that the data collected from both first-party and third-party cookies (used in both the JavaScript and the log file method) is frequently stored and analyzed by independent companies. Some companies aggregate this data among many sites and report broad ecommerce trends to media and other sources.”

Says Marshall, “Both the JavaScript method and the log file method provide valuable analytics data. It’s like choosing between a gas-powered vehicle and a diesel-powered vehicle. Each has its place, and an owner must decide which vehicle fits his needs.”

Kerry Murdock
Kerry Murdock
Bio   •   RSS Feed