A Tower of User Analysis Power

Before visiting TowerRecords.com, the web site of the big music retail chain, music fans often go to such competitors as CDnow.com, MSN.com and Amazon.com.

Before visiting TowerRecords.com, the web site of the big music retail chain, music fans often go to such competitors as CDnow.com, MSN.com and Amazon.com. These visits indicate the shopping-around nature of an online music transaction. The competing sites are previously visited by almost 20 percent of the people who later visit TowerRecords, according to statistics from end-user tracking service WebHancer Corp.

But even more significant, the single biggest source of visitors is Audiogalaxy.com, a music selling and music swapping site. While Yahoo.com, CDnow and MSN constitute 6.8 percent, 6.7 percent and 6.6 percent, respectively, of the sites previously visited by TowerRecords visitors, Audiogalaxy sends TowerRecords 9.6 percent of TowerRecords visitors, according to statistics published on end-user tracking site Realenduser.com.

Why are people who use the music swap programs, such as Napster, Kazaa and Morpheus, going from a music sharing site to a music purchasing site? Do the Audiogalaxy users come to buy, or do they come to scout the names of tracks by singers that they like and then seek a free download by its correct title? Ottawa-based Realenduser.com hedges on the answer.

"This could be an indication users are surfing Audiogalaxy.com only to find that the songs they want are unavailable, and then they are turning to TowerRecords.com to find and purchase music. Alternatively, users may be navigating to TowerRecords to find out the names of the tracks on a CD in order to have more specific search criteria while downloading music," says a Realenduser report.

The answer could lie in the customer click-stream data that can be captured on the TowerRecords site. TowerRecords officials, in West Sacramento, Calif., declined to comment by press time on the data posted at Realenduser.

The data was collected without TowerRecords participation and aired on Realenduser, a site sponsored by WebHancer to show off its end-user data collection capabilities.

Nevertheless, knowing that the largest percentage of visitors coming to TowerRecords were from an MP3 swap site could change some of the things TowerRecords does on its site. There might be a way to appeal to the swap-oriented set with new releases not yet generally available or special deals on music popular with this customer segment.

To find these avenues to profit, TowerRecords would have to make use of Web site analytics, looking at enough Audiogalaxy users to see what behavior patterns and motivations they bring to the TowerRecords site. Again, TowerRecords officials declined to comment.

The click-stream analysis of Web site analytics is worth a lot more if you know where your customers have come from and where they go after leaving your site, said Stephanie Bigusiak, WebHancers vice president of product management, in Ottawa. Click-stream data tends to be collected by a business on its own site. End-user data, collected from end users own machines, tracks behavior across sites and is much harder to acquire.

Knowing several previous sites that your customers clicked through before arriving at your site and knowing what they do as they move on to other sites is beyond the reach of most Web analytics software, also referred to as customer analytics in the customer relationship management field.

The statistics shown on Realenduser.com, on the other hand, indicate visitors to TowerRecords not only most often come from Audiogalaxy.com, but they most often leave TowerRecords for Audiogalaxy.com as well (6.6 percent). They also go to CDnow.com (2.43 percent), Yahoo.com (2.39 percent), Amazon.com (1.67 percent), Google.com (1.63 percent) and MSN.com (1.1 percent), indicating they leave for competitive shopping sites, the Realenduser report said.