There are some great nuggets of information in hard to find corners of the Internet.

In Noam Cohen’s recent blog post on the death of Encarta, Tom Corddry who “ran the team that created Encarta” has some reflections on the product. Instead of hiding in the comments section, I’m reprinting it here.

I ran the team that created Encarta, so I’m standing up to say a few awkward words at its graveside memorial service. Encarta, may it rest in peace, deserves to be remembered more for its quality than you suggest. Your sources repeat several notions that were never true of Encarta-first, that the content from Funk and Wagnall’s was “low quality” compared to Britannica, and second that the value added by Microsoft was primarily “graphics and sound.” The text from Funk and Wagnall’s was far superior to Britannica’s as a starting point for a digital encyclopedia, because it was much more nearly “structured data,” meaning that the architecture of the text was very consistent from one article to the next. This allowed us to add a lot of “contextual” value–to compute the relatedness of every article to every other article, and build what was at the time a uniquely useful set of links and navigational tools across the entire content. Britannica, by contrast, was a bloated mishmash, a consequence of its long tradition of having articles written by many different celebrity authors. (I ghost-wrote one myself, in fact). By the standards of the print encyclopedia world, Microsoft invested heavily in expanding and updating the content of Encarta right from the beginning. We consciously invested in the contextual value just described, in expanding the core content, in creating the world’s first truly global encyclopedia, and in an efficient update cycle. We had enough “multimedia” in the original product to keep the reviewers happy, but focused on the overall usefulness of the whole product much more than on the relative handful of video clips, etc. I’d argue that within its first five years, Encarta became the best encyclopedia in history: it had tremendously consistent quality and usefulness across a very broad range of topics, and added a great deal of value by the relationships it illuminated between topics. All of that has been rendered a bit quaint now, but in it’s day it was an accomplishment worthy of a graveside toast. Encarta had more than “the potential” to unsettle the print encyclopedia business–it pretty much destroyed it. Print encyclopedias were dead, thanks to Encarta, before Wikipedia existed. We expected from the beginning that Encarta would eventually be superceded by online information-seeking. As brilliant as Wikipedia is, I don’t think that Wikipedia by itself killed Encarta. I think the Web as a whole made Encarta obsolete. I hope treasured old copies of Encarta will live on for a while in remote corners of the world, where people have scattered access to computers but little or no connection to the Web–school libraries in Africa, for example. In those places, even out-of-date copies of African Encarta, the only Encyclopedia of Africa ever published, will live on, and Joe Biden will forever be newly-elected. I’ll drink to that.

— Tom Corddry

UPDATE: Corddry’s second post also provides some insight on his time with Encarta.  Thanks to Sage Ross for the tip.

In response to Randonneur, post 15, a bit of explanation: Print encyclopedia editorial groups, even in their heyday, were actually quite small, and much of their work from year to year was devoted to removing content in order to make room for other content. The size of the multi-volume sets was fixed, so every word added had to be offset by a word subtracted. Since it was also expensive to touch more pages than necessary when making changes (a printing fact of life), the editors ingeniously found ways to remove content as close as possible to where they were adding content. Need a big new article on Bosnia? Better find stuff to cut from the articles about Bosporus or Boss Tweed. The senior editors at these publications estimated that at least half of the total editorial effort was devoted to this sort of non-value-adding work. At Encarta, by contrast, nearly all the editorial work added value–writing new articles, updating, expanding and improving existing articles, and, of course, adding the sorts of elements that computers could support that were truly valuable: the sound of a bassoon, the way gravity works in orbital models, and so forth. At its peak, the Encarta editorial staff was roughly four times the size of any of its print competitors, included many of the best people from those competitors, and was devoting much more effort to actual new and better content. Then there’s the whole international aspect… all print encyclopedias were highly nation-specific. Encarta was always global. In practice, this meant a core of universal content with “extensions” in each national area, and global licensing of content, which further increased the value created by the editorial staff. This model also works brilliantly for Wikipedia. By exploiting its advantages, The Encarta team, over a period of a decade (late 80s to late 90s), created a body of content that offered greater scope AND depth than its print competitors, then tossed in the advantages of navigation, multimedia, integrated updates, and low price. As a result, more copies of Encarta were sold, by far, in its 16-year run than were sold of all print encyclopedias combined in their several-century run. It reached many more school children world wide than any encylopedia had ever done before it. Wikipedia has since expanded greatly on that reach, and is a far superior resource, as long as you recognize the inherent uncertainty about accuracy–but even that is a useful lesson for life: there’s no such thing as ultimate authority. My reference to “treasured copies” doesn’t expect too much sentiment in the first world, just continuing usefulness in the third world, where computers are spreading faster than the Internet, and even an outdated copy of Encarta may be the best source of accurate information in the whole village.

— Tom Corddry