With Data Analysis, Less Isnt More

Companies don't know what they're missing with Web analytics.

One of the major concerns in technology nowadays is that too much data is being gathered about us—that the government, corporate America and various unscrupulous persons are using powerful data-analysis tools to invade our privacy.

This is a valid concern, and many eWEEK columnists have written about the need to staunchly and vigilantly protect personal privacy and the privacy of customers and business partners.

However, there are times when picking apart data is the right thing to do. But as I wrote in my review of Web analysis tools this week, many organizations arent effectively analyzing the data they do have to improve their business.

Whats more troubling, perhaps, is many companies that purchase powerful analysis and business-intelligence tools dont use them effectively. Users of these products often generate the most basic and obvious reports and never get their hands dirty with the deep-analysis tools.

By ignoring these products deep-analysis capabilities, organizations could be missing important trends—information that might show, for example, where a company is losing business.

Companies might also see that business decisions that were made based on basic information were wrong and ended up costing money in the long run.

People reading this column who are knowledgeable about analysis and reporting might be thinking that Im stating the obvious—that business decisions should not be made based on simple reports and executive-summary analysis.

But at a lot of companies, the people using data-analysis tools to make business decisions dont know a whole lot about statistical analysis and reporting. And theres a certain seductive rightness to the simple and obvious conclusion—it seems to make sense, and its easy to get higher-up executives to buy into your decisions using this data.

But, in many cases, that data can be flat wrong.

Say you manage a Web site, and you start to notice that certain types of Web content are getting a lot of traffic. You look at this data and think, "Wow, people really like this type of content; we should have more of it." So you redesign your site to include more of the content and to highlight it prominently.

But what if you had dug just a little deeper?

Maybe you would have noticed that most of the traffic you were seeing was coming from a reference at an external site. Maybe you would have noticed that the vast majority of the visitors coming in based on this reference werent clicking on any other site content. And maybe you would have noticed that your best site visitors—those who come to the site on their own and visit on a regular basis—didnt have that much interest in or affinity for the content that was drawing lots of traffic from "outside" users.

Based on top-line information, the site manager might redesign the site to attract visitors who have little or no interest in the core content of the site. In the process, the site manager discourages the best and most regular site users, losing traffic in the long run.

Based on deeper analysis, the site manager could make choices that maintain the attention of loyal and regular visitors. The site manager might also come up with a way to highlight the content thats attracting the outside group, without the site being perceived as a "me, too" destination when compared with the site that did the initial referral.

Analysis and reporting tools can be valuable allies in maintaining and building your business. By using these tools effectively and to the full extent of their capabilities, your company can gain valuable insight into customer and company activity, and build on your successes while avoiding repetition of mistakes.

But if you arent going to use these products properly—if youre just going to rely on the most basic information they can give you—you might as well stop spending good money on them.

Labs Director Jim Rapoza can be reached at jim_rapoza@ziffdavis.com.