Predictions for 1999 Revisited

by Jakob Nielsen on December 27, 1998

Sidebar to Jakob Nielsen 's column on predictions for the Web in 1999 .

In December 1998, I predicted six trends for the Web in 1999:

  • The Web has slowed down, and most important changes will take more than a year
  • Mobile access to the Internet
  • Resurgence of Web standards
  • Automated customer service
  • Web patent bonanza
  • Beginning Y2K problems in December

Now (writing at the end of 1999), we can revisit these predictions in light of what actually happened .

Slower Change: Yes

Definitely. The Web at the end of the year is almost the same as it was in the beginning of the year. Bigger, yes (10 million sites instead of 4 million), but most things are done about the same as they were a year ago.

The technology did not change much, the bandwidth did not increase significantly, the business models are about the same, and the main services are about the same.

Several important new sites were launched during the year, so it's not as if change has stopped: in particular, reputation managers are finally happening. It's just not as fast or dramatic as the explosion of ideas in 1993 and 1994. Back then, we felt the earth move every second week.

The bigger Web does have one interesting implication: in 1998, the default case was that something you wanted would not be available on the Web. In 2000, the default case is that anything you want is available on the Web - if only you could find it :-(

Mobile Access: Not Yet

I am very disappointed in the slow growth in wireless Internet access during 1999. In the United States, we were the victim of fragmented cellular networks and spotty digital coverage. And all over the world, we were victims of telephone companies that continue to try to sell cellular transmissions by the minute instead of by the packet.

Resurgence of Standards: Yes

It is now very rare to see a website that aims at bleeding edge design at the expense of Web standards.

Automated Customer Service: Didn't Happen

There were endlessly many articles written about bad customer service on the Web, but amazingly, most analysts still seem to think that call centers with human operators are the solution.

It would be much better to design sites that people could actually use and where they could find the solutions and answers themselves.

For example, if you review 100 random customer service emails (an exercise that is highly recommended), you will probably find a large bunch of the form "where is the product I ordered several weeks ago?" Most such questions can be avoided by

  • honestly disclosing what products are out of stock before the user has placed the order
  • informative email messages when orders ship or when they are backordered
  • a simple order tracking user interface

Patent Bonanza: Yes

Big time yes. One of the biggest stories during the year was's attempt to monopolize one-click ordering. Many companies bitterly regret that they did not pay more attention to getting patents.

Strangely enough there are still many start-up companies that do not file for patents on everything they invent.

Beginning Y2K Problems: Didn't Happen

So far (as of December 26, 1999), there have not been any serious Y2K problems on the Web. I still expect a few in January 2000.

Share this article: Twitter | LinkedIn | Google+ | Email