There is no source of reference remotely as authoritative as the Encyclopaedia Britannica. There is no brand as venerable and as veteran as this mammoth labour of knowledge and ideas established in 1768. There is no better value for money. And, after a few sputters and bugs, it now comes in all shapes and sizes, including two CD-ROM versions (standard and deluxe) and an appealing and reader-friendly web site. So, why does it always appear to be on the brink of extinction?

The Britannica provides for an interesting study of the changing fortunes (and formats) of vendors of reference. As late as a decade ago, it was still selling in a leather- imitation bound set of 32 volumes. As print encyclopaedias went, it was a daring innovator and a pioneer of hyperlinked- like textual design. It sported a subject index, a lexical part and an alphabetically arranged series of in-depth essays authored by the best in every field of human erudition.

When the CD-ROM erupted on the scene, the Britannica mismanaged the transition. As late as 1997, it was still selling a sordid text-only compact disc which included a part of the encyclopaedia. Only in 1998, did the Britannica switch to multimedia and added tables and graphs to the CD. Video and sound were to make their appearance even later. This error in trend analysis left the field wide open to the likes of Encarta and Grolier. The Britannica failed to grasp the irreversible shift from cumbersome print volumes to slender and freely searchable CD-ROMs. Reference was going digital and the Britannica's sales plummeted.

The Britannica was also late to cash on the web revolution - but, when it did, it became a world leader overnight. Its unbeatable brand was a decisive factor. A failed experiment with an annoying subscription model gave way to unrestricted access to the full contents of the Encyclopaedia and much more besides: specially commissioned articles, fora, an annotated internet guide, news in context, downloads and shopping. The site enjoys healthy traffic and the Britannica's CD-ROM interacts synergistically with its contents (through hyperlinks).

Yet, recently, the Britannica had to fire hundreds of workers (in its web division) and a return to a pay-for-content model is contemplated. What went wrong again? Internet advertising did. The Britannica's revenue model was based on monetizing eyeballs, to use a faddish refrain. When the perpetuum mobile of "advertisers pay for content and users get it free" crumbled - the Britannica found itself in familiar dire straits.

Is there a lesson to be learned from this arduous and convoluted tale? Are works of reference not self-supporting regardless of the revenue model (subscription, ad-based, print, CD-ROM)? This might well be the case.

Classic works of reference - from Diderot to the Encarta - offered a series of advantages to their users:

1. Authority - Works of reference are authored by experts in their fields and peer-reviewed. This ensures both objectivity and accuracy.

2. Accessibility - Huge amounts of material were assembled under one "roof". This abolished the need to scour numerous sources of variable quality to obtain the data one needed.

3. Organization - This pile of knowledge was organized in a convenient and recognizable manner (alphabetically or by subject)

Moreover, authoring an encyclopaedia was such a daunting and expensive task that only states, academic institutions, or well-funded businesses were able to produce them. At any given period there was a dearth of reliable encyclopaedias, which exercised a monopoly on the dissemination of knowledge. Competitors were few and far between. The price of these tomes was, therefore, always exorbitant but people paid it to secure education for their children and a fount of knowledge at home. Hence the long gone phenomenon of "door to door encyclopaedia salesmen" and instalment plans.

Yet, all these advantages were eroded to fine dust by the Internet. The web offers a plethora of highly authoritative information authored and released by the leading names in every field of human knowledge and endeavour. The Internet, is, in effect, an encyclopaedia - far more detailed, far more authoritative, and far more comprehensive that any encyclopaedia can ever hope to be. The web is also fully accessible and fully searchable. What it lacks in organization it compensates in breadth and depth and recently emergent subject portals (directories such as Yahoo! or The Open Directory) have become the indices of the Internet. The aforementioned anti-competition barriers to entry are gone: web publishing is cheap and immediate. Technologies such as web communities, chat, and e-mail enable massive collaborative efforts. And, most important, the bulk of the Internet is free. Users pay only the communication costs.

The long-heralded transition from free content to fee-based information may revive the fortunes of online reference vendors. But as long as the Internet - with its 2,000,000,000 (!) visible pages (and 5 times as many pages in its databases) - is free, encyclopaedias have little by way of a competitive advantage.