Analyzing Web customer patterns

Read how Jubii used SPSS' CustomerCentric softwarer to analyze Web traffic

ABOUT THE VENDOR SPSS Inc. is a worldwide provider of analytical technology for business, government, and higher...

education. The company's products enable customers to make better decisions by learning from the past, understanding the present, and anticipating the future. With this insight, organizations gain a true competitive advantage: the ability to manage the future. SPSS analytical technology is brought to the market through five divisions: CustomerCentric Solutions (for integrated analytical CRM solutions); SPSS BI (for data mining and statistical products and services used to solve business problems); ShowCase (for analytical products operating on IBM iSeries/AS400 platform); SPSS MR (for analytical solutions in the market research industry); and SPSS Enabling Technologies (for licensing SPSS technologies for use in other analytical applications). ABOUT THE CLIENT A subsidiary of Lycos Europe, Jubii ( is Denmark's most popular Internet portal, with 2.3 million visitors per month. Because Jubii's services are free, its commercial success depends on earnings from banner ads, sponsorships, Net events, newsletter advertising and similar activities. Such a business model makes it all the more important for the company to know exactly what its users want to buy and what information they want to see -- in other words, to be customer-centric. ABOUT THE TECHNOLOGY CustomerCentric is an analytical CRM solution that enables companies to use packaged predictive models to minimize risk as well as maximize revenue, ROI, and Lifetime Value. For companies preferring to develop their own predictive models, SPSS offers an array of data mining and analysis products including Clementine, a data mining workbench. Jubii leveraged SPSS' analytical power to study patterns of Web visitor behavior and develop customer profiles to optimize the content and placement of ad banners on its Web sites. SearchCRM spoke with Kasper Larsen, Jubii's sales director and partner, about the project. -------------------------------------------------------- SearchCRM: Why did you decide to look at CRM solutions? Larsen: Because we know that there's so much valuable information about our customers lying around which we didn't use before. It's a question of getting access and finding out the right way to use that information to derive even more customer satisfaction in the future. SearchCRM: Who are your customers? Larsen: There are two groups of customers: The free riders--the ones using our Internet pages--and the banner customers, the advertising customers who pay to get promoted to the free riders. We need to optimize information to make our paying clients happier and our non-paying clients less bothered by the advertising. SearchCRM: What were you looking for in a CRM vendor? Larsen: The big question was, how much traffic can they actually handle? The amount of data we have is enormous compared to other types of companies. A grocery store, for example, has a transaction and a couple of products per person per week. We have 700 or 800 transactions per day per customer on average. The data load is much bigger than most CRM providers are used to. We were also looking for a big name brand because it was a huge task and very few people have come out with a solution that would actually be what they thought it would be. We wanted a big organization so we weren't dependent on one or two people. SearchCRM: What type of CRM were you using before SPSS? Larsen: We used a system to do part of the task that was put together by four different suppliers. There was constantly a fight over whose fault is this and whose fault is that that the system isn't working. It ended up that the company pulled their product off the market and gave us our money back because they realized that having four companies fighting among themselves wasn't satisfactory for the customers. We learned a lot from that lesson. SearchCRM: What was the most challenging aspect of the technology implementation? Larsen: Figuring out which way to analyze the data. There are so many combinations to analyze the data -- some are good for some things, and others are good for other things. It's often a trade-off of being able to mine down to the individual or being able to get a simplistic overview of groups of customers. I think we did five to seven trials before we figured out what was the right combination for us. But once we decided on the where and when and how, the implementation was quite easy. It was fast and there were almost no problems. The only trouble was a bit of a resource problem where suddenly it was vacation time and everyone forgot to put that in the product schedule. SearchCRM: How do you use the system in your day to day operations? Larsen: The system is a plug-in to DoubleClick. We host and maintain a DoubleClick ad server. When people visit our registration page, we write some special cookies that will be eaten by the ad server and passed on to the optimization system. The ad optimization system shows a banner ad only to the people that have shown prior interest to that type of banner. We can pick out any campaign and ask the system to optimize it to get optimum response. SearchCRM: What have the results been so far? Larsen: We've seen more than a 20 percent average increase in response. SearchCRM: Do you have any advice for companies that are starting a similar project? Larsen: Go out and find a supplier that you feel you can work with before you start the project. We're talking about a lot of trust -- in the end, it's very much about trust. Who do you trust to make sure that your companies survives into the future? Linda Formichelli's writing appeared this year in Woman's Day, Wired, Writer's Digest, Family Circle, Psychology Today. Contact her at, or check out her Web site

This was first published in September 2001

Dig Deeper on CRM analytics and business intelligence



Find more PRO+ content and other member only offers, here.



Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to: