Alertbox 5 Years Retrospective

by Jakob Nielsen on May 28, 2000

The first Alertbox was dated June 1995 and published May 25, 1995 (one of the few times I have been ahead of deadline :-)

In 1995, I published 7 Alertboxes which generated 30,000 page views: only 4,300 per column. Rather disappointing; I almost gave up. Now, in mid-2000, the Alertbox gets about 210,000 page views per column.

Thus, readership grew by 4,800% during this 5-year period.

On the other hand, the Web had 35,000 websites when I started publishing the Alertbox; it has 16 million sites today (a growth of 45,600%). In other words, I am losing ground in proportion of the Web covered by the usability message.

The single-most popular column, top-10 mistakes of Web design from 1996, currently gets 40,000 page views per month and is getting close to having been read one million times during the four years it has been online. Even this is a losing battle: if we simplistically assume one designer per site and one reader per page view, the numbers imply that 15 million websites have been designed without the benefit of knowing these ten basic rules.

Basically, though, the traffic growth proves the value of publishing a column that dares take a strong stand and has a singular voice. Stick to it, even when the initial results are disappointing. Growth snowballs on the Web, as more and more sites link to the column (I currently have about 20,000 inbound links from other sites). Of course, to pick up these links and the associated growth, it is important to make the original URLs live forever (most publishers cluelessly let old columns suffer linkrot).

Alertbox Victories

The biggest victory for the Alertbox may be that almost everything I have written in 105 columns over five years remains relevant today. For example, the guidelines in my column on Web-based multi-media from 1995 are still correct. Very few people who write about the Internet can say the same; no big companies dare put their five-year-old reports on the Web for potential customers to check the sustainability of their predictions and advice.

The reason old Alertboxes continue to offer valid advice is that they are based on human characteristics and not on the details of the technology or the latest tradeshow buzzwords. People change slowly. In fact, to shape a profitable Internet strategy, you are probably better off running a usability test with five customers than reading the latest volume from the Harvard Business School Press.

The Alertbox was founded on the principle that success on the Web comes from simple rather than glitzy design and from making it easy for users to do what they want to do rather than imposing your own internal perspective on the users.

Simplicity has won: it is rare to see major sites launch with a flashy design since it is known to lead to disaster. (One of the few exceptions was Boo which went bankrupt  this month - almost like a birthday present to the Alertbox.) My analysis of the "top-ten" mistakes three years later showed that these classic mistakes are much less frequent these days. Furthermore, the biggest sites commit the fewest of the usability mistakes (that's how they get to be big).

Other correct calls:

Alertbox Defeats

Even though most of my predictions have come true and most of my recommendations have proven their value as the foundation for Internet success, the record is certainly not perfect.

Most notably, I frequently predicted the emergence of micropayments as one of the main business models for the Web. Still waiting. I repeat my prediction from 1998: micropayments will happen in about two years (of course, this now means 2002, not 2000). I remain convinced that the only way to ensure high-quality Web content is to have the users be the paying customers. We are currently passing through a temporary phase of the Web where everything has to be free. At some point of time, people will tire of low quality and ever-more-intrusive ads and start paying.

In August 1995, I outlined five generations of online services, with Mosaic being the third generation. I predicted the emergence of a fourth generation based on examples like AT&T Interchange, MSN, and Java. This did not happen - instead AT&T Interchange died due to lack of Internet integration and MSN withered due to clueless attempts to turn the Internet into a television metaphor. Users have suffered Web browsers long enough, and it currently seems like we may jump directly to the fifth generation of Internet interfaces and bypass my "fourth generation" completely.

The miserable quality of Web content also counts as a major defeat for the Alertbox. I have repeatedly campaigned for a new writing style that is optimized for the Web. Despite a 1997 Alertbox showing that a website can double its usability by following the writing guidelines, virtually all sites still rely on old-media writing:

Neglected Columns

Old columns that are important but attract low traffic:


Alertbox Ten Years (2005)

See also: Reader comments on this Alertbox (including why the Alertbox reaches further than indicated by page view stats and why WAP may provide the first micropayments).

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