How ONIX Came To Be

In 1997, superstore behemoth Barnes & Noble had launched their website in competition with Amazon. (As a matter of disclosure: I began working there in 1998, directing the database that served both the web and the stores.) Borders, Hastings, Books-A-Million, and others soon followed. There were numerous start-ups dedicated to selling specific sorts of books – Varsity Books for textbooks, FatBrain for business books. Companies were acquired, rose up, shut down – it was absolutely chaotic.

In the midst of all this, the problems presented by back-office, transactional metadata (truncated titles, metadata in ALL CAPS) were abundantly clear to consumers – these websites were ugly, clunky, and not very enticing. Publishers noticed. Distributors noticed. Everyone saw an opportunity to increase sales.

ONIX stands for Online Information Exchange, and was developed as a joint effort by the Association of American Publishers (AAP) and EDItEUR (which originally stood for EDI-to-Europe, but which has evolved more broadly into a London-based standards body for the book industry). It was created to solve two problems: (1) that consumers were now looking at this data so it had to be more robust, descriptive, accurate, and reflective of what they needed to see, and (2) that ANSI X12, as a US-based standard, was insufficient for international communications about books.

I was on the front lines back then – it was hotly competitive between Amazon, Borders, and Barnes & Noble. There were lawsuits, front-page news articles, insults, and shade thrown. It was nasty. But everybody could agree that the metadata was causing us all the same headaches. So there was a parley.

In 1998, in the conference room at AAP on 5th Avenue, the Big Seven publishers (yes, there were seven at the time), Barnes & Noble, Amazon, Ingram, Baker & Taylor, a handful of startups, and a number of other interested companies sat down at a large conference table and laid out the problems. It was the first time B&N and Amazon had allowed representatives to sit in the same room together – Cindy Cunningham and myself. (We later became good friends – largely due to this experience.) It was clear that we needed to present a unified front to persuade publishers to adopt this new standard that would benefit all of us. After two years of negotiations, ONIX 1.0 was published and its maintenance in the US was handed to the Book Industry Study Group, which created a metadata committee (now run by Richard Stark at B&N, who was one of my first hires there) to handle changes and fixes and additions to the code lists.

ONIX is not terribly sexy. But it allowed sexiness to happen. As we evolve through ONIX 3.0 and beyond, knowing how we got here will help us lay the infrastructure for whatever awaits the book industry as the Web itself advances.