Tracking a Trend, Part II

A realization that the ability to understand Web site behavior has a direct effect on customer perception and business success is driving the growth rate of the Web analytics market.
By Lauren Gibbons Paul

(A new generation of Web-based business intelligence tools can do much more than count and collate page view information. Today, in Part II, contributing writer Lauren Gibbons Paul says that Web BI are being used to hone business strategies, target marketing promotions, ease site navigation, and more.)

The Web analytics market is growing at a rate of 200 percent annually, projected to reach $4 billion in 2004 from a 1999 base of $141 million, according to a report from The Aberdeen Group, of Cambridge, Mass. Driving this healthy growth rate: the realization that the ability to understand Web site behavior has a direct effect on customer perception and business success. According to Guy Creese, research director, Internet analytics, for the Boston-based Aberdeen Group and author of a report, "Web Analytics: Translating Clicks into Business," IT managers are using Web analytics to forecast Web site usage and demand; segment customers and target them with specific offers; find alternative ways to understand customer behavior; and fulfill expectations carried over from the brick-and-mortar world.

One example: Creese had as a customer a music/CD site. Company executives believed their target audience was teenagers with disposable income. But using Web analytics data, it turned out that the site's second biggest customer group was adults, ages 55 to 75. "It was grandparents trying to buy music for their grandchildren. They were hungry for advice on what to buy," says Creese. Armed with this insight, the company devised content and promotions specially targeting this group.

But those who have done the implementation say it can be complex to get these tools or services up and running. Although WebSideStory hosts its HitBox Enterprise application for ABC Distributing Inc., a North Miami, Fla., catalog retailer, it took ABC's Web site developer Daniel Gudema two weeks of full-time work to customize the product so it would be usable for ABC. In fact, adds David Green, ABC's Web site manager, anyone who wants to use this type of product would do well to sit down with a blank sheet of paper and sketch out an "information architecture." That is, how the information should be structured so it is more useful for the users. Green and Gudema met with every functional group that would be using the service, from the executive boardroom to marketing to customer service, to see what kind of information that user might like to see. Also of critical importance: defining the keywords the users would use to search the system.

"You have to do a fair amount of work establishing the naming conventions to be used within the tool, and then you have to stick with them," says Gudema. He learned that lesson the hard way. "In the first month, I just stuck any old term in there." Things can get out of control quickly when there are one or more terms that describe the same thing, he says.

Getting Inside the Customer's Head

Jonathan Segal, now an IT consultant, used WebCriteria, a customer experience analysis application from WebCriteria Inc., in 2000, while he was the manager of interactive media at Verizon Communications (then GTE) in New York. WebCriteria uses an electronic agent called Max to act as a typical customer would on a Web site. When a company signs up for the customer experience analysis service at the WebCriteria Web site, Max visits the company's Web site and surfs as a typical visitor would, based on an algorithm. The user can then access the customer experience data on WebCriteria site. "Max offers a very different perspective on the user experience. He gives you an idea of how easy it is to navigate from one area to another. He'll provide problem pages, broken links, pages no one ever goes to," says Segal.

Lessons Learned

Lessons Learned about Web Analytics (specifically HitBox Enterprise)

--Create an "information architecture" before you do anything else. Interview the would-be users to see what information they need from the system. Then be sure you have the categories, keywords and terms defined and structured before you go ahead with implementation.

--Don't overestimate the ease of use. Executives who have never used a decision-support or CRM tool before are likely to feel overwhelmed by the amount of information an application such as HitBox Enterprise supplies. Help them home in on the data that is likely to be more useful for them.

--Understand your tool's terminology. Before you go telling the business executives your site had 1,000,000 hits yesterday, you'd better get straight exactly what this means in the context of your Web analytics application. (1,000,000 hits might translate to 100 unique visitors.)

The challenge came in interpreting the monthly reports--which included link path and broken link data--and translating that to action. Since Segal was responsible for improving the customer's Web experience, he used the data to improve the navigation on the site and to prove to his colleagues that the search engine needed to be overhauled. Marketing personnel would use the data to plan new service offerings targeting specific groups of visitors. The payoff: GTE's site was named the most usable b2b Web site in Business 2.0 Magazine in 2000. The site offers a full range of GTE services and products: Customers can order phone books, buy cell phones, and sign up for Caller ID service, among other things.

The sheer depth and breadth of data available from Web analytics tools can be daunting. Jim McCarthy, director of marketing for Kiko Inc., a knowledge network company in Long Beach, Calif., uses Web Business Analyzer from MicroStrategy to track activity on Kiko's Web site. "I could do a report on Web Business Analyzer that would tell me all of the people in Hawaii over the age of 35 who uploaded a file last week. And it's impressive that the tool can do that. But our metrics are pretty simple and straightforward: How many people joined the site last week and what did they do?" says McCarthy.

Another trend: Companies are scrutinizing Web traffic information to determine the profitability of a particular path a visitor takes through the site. For example, one shopper on a clothing site might click on a "Trends" link highlighting the most fashionable items and then make a purchase, while another shopper might click on the "Sweaters" link and then buy the item. "You might find the 19th most popular path is five times more profitable than the most popular path," says Creese. Then you can re-vamp the site navigation to try to steer traffic down the more lucrative path.

No matter what type of Web analytics application you choose be prepared for a surge in interest from the users, who tend to pore over data they never had access to before. Says ABC's Green, "Our executives were a little intimidated by the tool at first. Now, they're beginning to ask for ways to see more and more information."

Freelance writer Lauren Gibbons Paul covers business and technology from her home in Waban, Mass. Reach her at laurenpaul@mediaone.net.






0 Comments (click to add your comment)
Comment and Contribute

 


(Maximum characters: 1200). You have characters left.