According to the wired article, some of the reasons have to do with complexities related to going online. In order to resolve those issues, the industry needs to play closer attention to user experience and start figuring out how to make things easier. Return on investments in technology will increase if more people use a system. More people will use a system if it’s easier to use. However, few companies pay close attention to those kinds of details. Next time someone asks you why usability research is needed, point out the relationship between usability and bottom dollars: the business people will immediately see the value.
A question remains, however, on how to get people back. As standard marketing theory often points out, it is easier to convert a customer that has never used a product than to get one that has unsuccessfully used a product to come back. This is a challenge that marketers everywhere need to crack in order to increase overall market shares.
Also of interest in the story is the fact that most disabled people do not go online. This seems to represents a huge market for anyone as the Internet could actually act as a great enabler for disabled people. Site developers should pay close attention to things like the World Wide Web Consortium WAI project which could help lure more disabled people online.
Last but not least is the question of economics and Internet access. Obviously, there is a need for education here as most libraries now offer Internet access for free. However, recent efforts in monitoring what is done online in libraries probably keeps some people from getting online and the remainder is probably either not aware that Internet access is available at their local library, not interested in going to the library (for educational reasons or other), or doesn’t see any value in the Internet. A way to solve this would be through better educational programs related to the fact that the service is available, addressing quick wins by showing people how they can do certain things faster online (for example, showing how people could get a better job or reduce the amount of time and money it takes to do something by going online).
However, all of these points have a pre-requisite. Before addressing the problem, we must first understand who those people are. A demographic make-up of Internet drop-outs could help (are those people mostly from a certain age group, for example) in understanding whether this might become a longer term trend that needs to be addressed or whether most of the problem is going to go away over time (as kids nowadays seem to get online at a faster rate than older people).