|
|
holds web's real hope
Custom content to serve users and advertisers
By Andy Wang
The promise of advertisers flocking to the internet because of highly targeted marketing opportunities has largely been unfulfilled.
But while many mainstream advertisers might be forsaking the web right now, the promise of so-called "data mining" could bring a new slew of them online.
Although few people realize it, web sites are about to rely on user data to get much smarter and create a more personalized experience for each user--whether visitors want it or not.
These smarter sites will create what is arguably more compelling content for users and more compelling marketing opportunities for advertisers.
This will all be somewhat intrusive, a little like feeling that Big Brother knows too much. It also, however, could be quite cool. The user experience will be more efficient, as will the advertising opportunities.
Last week it was reported that the New York Times Digital plans to offer new subscription-based content to its 15 million registered users. While much of the content on Times Digital properties like NYTimes.com and Boston.com will remain free, the company plans on using the demographic data of its registered users to help create premium services.
It's about time that the Times attempted something like this, and this is something other web sites should be watching closely. After all, many sites have user data that hasn't been fully utilized.
What's more, data mining can be used for much more than creating subscription-based content.
The idea is fairly simple in theory. A web site or an online advertising company can study the so-called "click-streams" of its visitors to understand user habits with extraordinary specificity. They can find out what sections and pages of a site an individual user likes to visit and then dynamically generate a site to cater to that user. They can determine what kind of content a user might like and what kind of content doesn't seem to interest the user at all.
Data mining can be employed to tailor a web site to every single individual user, creating a more efficient surfing and reading experience. And with this user data in hand, sites will be able to attract advertisers who are looking to target a precise audience.
That is to say, if a user visits a site 50 times and never once goes to the sports section, the link to the sports section will be made much less prominent while the real estate it used to occupy will feature a link to content or advertising that the user is more likely to be interested in.
At a South by Southwest Interactive Festival panel in March, Slate publisher Scott Moore said that his site was already looking into data-mining possibilities.
This sort of data mining could also save struggling online advertising companies like DoubleClick and Engage. After all, advertising campaigns that are more highly targeted should command higher rates.
Now you can expect privacy advocates to have a fit about all this, but they'll just have to deal with reality.
In a day and age when lots of personal information is already easy to find online--a day and age when a busboy can use a library computer to hack into bank accounts of millionaires--the reality is that most people won't be bothered by the fact that a site knows they prefer the business section over the entertainment section, or vice versa.
What's more, if and when data mining is fully utilized, advertisers will greatly prefer highly targeted sites. Visitors, meanwhile, will also be more likely to pay for content that is based on their particular interests. This means, of course, that those who want to survive are going to have to make their sites smarter.
So bully for The New York Times Digital and its subscription plan. This particular experiment--like all current attempts to charge for content--might not work, but it could lead to something that will.
April 5, 2001 © 2001 Media Life
-Andy Wang is features editor for Gear magazine and founder of Ironminds.com
Printer-Friendly
Version | Send
to a Friend
Cover Page | Contact
Us
