In my January 1996 column on Trends for the Web in 1996 , I predicted
- the death of "surfing" as a dominant way of using the Web
- websites will give their users individualized service to build up relationships
Even though I don't know of official statistics, my qualitative data from user interviews and usability studies does tend to support the death of surfing: most users stick to a small number of favorite websites and only go elsewhere when they have specific tasks to accomplish. Most people don't have time to check out cool sites just for the sake of looking around on the Web.
Most customizations are quite simple, however, and don't really amount to building a relationship between the user and the website. It is still too complicated to program advanced Web services on the server, so most sites don't do so.
In January 1996, I highlighted several sites that took the initiative to contact their users when they had appropriate new information for them (for example, SunSolve notifying users of when a patch was available fix a bothersome bug and Amazon.com notifying users when their favorite author has published a new book). These services are quite similar to the so-called push model for Web content that became popular later in 1996 (link to New York Times story requires registration - good example of the unpleasant logins I referred to above, but at least the Times does save your userid in a cookie) .
Unfortunately, most of the "push" technologies simply flood the user with irrelevant content in the hope of capturing as many eyeball-minutes as possible. PointCast is a productivity sink that constantly tempts users with news that serve as the perfect procrastination device to prevent them from getting any work done. The push services I liked in January 1996 were highly selective and only contacted a user in those rare cases where something that was valuable to that particular user had happened.