Tuesday, May 24, 2005

Want To Know What To Read? Storycode.com Will Do the Math
May 20, 2005
By Rachel Deahl

The notion that taste is personal may have gone the way of eight tracks and Hammer pants. A number of new companies are pioneering a variation of the search engine, known as "recommendation technology," which would use hard data to essentially tell consumers what music, movies and clothing they will like . . . even if they don't know it yet themselves. While this software is already present in the book industry—anyone who's ever been startled by Amazon or B&N's seeming omniscience ("Geez, how'd they know I'd love Knitting to the Oldies?") can attest to this—some companies are now gearing up to bring it to readers in a more ambitious way.

Storycode, founded by Steve Johnston, is one such example. Currently in "soft launch" mode (as of March 1), Johnston claims his new venture—it's intended to function as a book-recommendations database for fiction titles, powered by a "coding" system—has the potential to change the face of the publishing industry. The codes, culled from readers' responses to a series of questions—queries are broken down into five categories and touch on areas like plot and characters—are used to classify, and compare, every book featured on the site. Adding that the site will not publicly launch until a critical mass of codes has been entered—this could mean waiting for books to receive upwards of 20 codes each—the system is intended to offer readers recommendations of books, based on statistical data. In other words, if you just finished The Corrections and loved it, and now want to read another novel just like it, Storycode, Johnston says, will provide a recommendation, complete with a percentage quantifying the similarity between it and a list of matching books.

If the notion of a quantifiable classification system for books, which are, after all, judged and experienced by readers in inherently personal ways, sounds dubious, Johnston counters that all fiction conforms to certain guidelines.

"Every novel does, in some combination, form to a classic story type," says Johnston, who previously spent his professional career in two places: behind the counter of a bookstore; and in front of a computer. After a decade as a bookseller, Johnston became a Web consultant, offering his services to business owners looking for ways to successfully leverage the power of the Internet. "Even with more sophisticated anti-plot novels, the codes will reflect the ambiguity of responses that they receive." He says he sees Storycode as "filling a hole in the retail book trade," and predicts a time when retailers could offer booksellers a more satisfying and viable way to guide their customers to more informed purchases—what he calls "genuine recommendations." The Storycode future he envisions has booksellers logging on to the site from behind their counters, or e-booksellers repurposing his database online. Independent booksellers, meanwhile, are the "potential evangelists" for Storycode.

Leonie Flynn, a former independent bookseller and an editor of The Ultimate Book Guide (A&C Black), says she thinks Storycode is a good idea but doubts booksellers will latch onto it the way Johnston predicts. Saying the site was "more a browsing tool than a fast, immediate selling tool," Smith foresees the site as a place for book lovers over industry professionals. Andrew McClellan, manager of books for Virgin Megastores, says right now Storycode is "no more than an interesting idea." McClellan says that Storycode has potential, but wonders if it can be fulfilled.

"If they receive a lot of input from members of the book reading public, then the site may provide a credible and organic resource of recommendations…as well as becoming a hub for book readers online," he says." From that point, retailers and publishers could make great use of it; however, that's a whole lot of ‘ifs and ‘mays.'"

The idea behind Storycode echoes what other technology companies have been doing for some years for other kinds of products. ChoiceStream, Inc., for example, a Mass.-based company that delivers recommendations across a number of fields, including music, general retail merchandise, television and movies. The company licenses its database to companies such as eMusic, AOL and Yahoo. Its software is programmed to not only know just what consumers like, but more importantly, why they like it.

Darren Gill, ChoiceStream vice president of business development for entertainment, says the company's database is essentially a more advanced version of the recommendations systems on sites like Amazon and Netflix (which offer lists of recommended books and DVDs to consumers based on items viewed, bought or rated). According to Gill, because ChoiceStream has seven million users in its network, with a massive index of products classified, the company has a more advanced system for delivering personalized recommendations.

"Storycode is asking you to do the background work," Gill says. "What we need from a user is to have them tell us about content; people who have done three or four ratings on Yahoo Movies, which licenses our database, get a personalized experience." Gill also says that ChoiceStream is looking to expand the reach of its business to cover other areas—with literature being one of those. "Books has been on our list for a while," he says.

Savage Beast Technologies, Inc., another company in the recommendations business, works exclusively with music and offers recommendations through a system it calls the Music Genome Project. The company, which currently licenses its technology to retailers like Best Buy and Borders, employs a team of specially trained musicians (the profile of a "music analyst" is someone with a four-year music degree and a background in music theory) to characterize songs according to some 400 different attributes. Tim Westergren, one of Savage Beast's founders, says he sees recommendation technologies, and the businesses that provide them, becoming more pervasive in our daily lives. "You can [apply this kind of system] to almost anything from food to dating. I mean, Match.com is trying to do it with people." Westergren says he believes "it's the next wave of the search."


No comments:

Search This Blog