ISBN Lookup – the Good, the Bad and the Ugly

Over the last few years the ISBN has become a lot more important to many, if not most, online booksellers. Many of the larger sites are using ISBN lookups to match a bookseller’s listing to information about that title pulled from an ISBN database. When it works, it’s a quick and efficient way to provide additional information about a book to supplement the bookseller’s listing or free the bookseller from having to enter that information themselves. However, often it results in matches to information about a different edition of the book or, in some cases, matches with a completely different title than the one being listed.

Many booksellers are understandably frustrated by this, and many are questioning the need and/or feasibility of an ISBN lookup function. But there is a method to the madness, even though in it’s current state it often does more harm than good. This article will examine how and when an ISBN lookup can be useful along with the pitfalls of using it today.

Buying Books Online

The market for used books had never been a homogeneous one. There have always been several different kinds of used books being sold for different reasons, from rare books sold to collectors to common mass market paperbacks sold to readers on a budget. There have also been different kinds of booksellers, from Antiquarian sellers who issued catalogues and whose store was open “by appointment only” to your local paperback exchange. No one would have expected these disparate sellers to use the same methods to sell their very different wares.

That seemed to change with the onset of online bookselling. In the beginning online bookselling was used in a way similar to the paper catalogue, with a limited but highly targeted audience. It operated in many ways like a giant catalogue of books from multiple booksellers that could be searched. But then the more general used books sites began to appear: Bibliofind and the Advanced Book Exchange appeared; Interloc morphed into Alibris; and Ebay arose and eventually merged. And there was… changed the nature of used bookselling. As one of the first online merchandisers (and arguably the most successful, at least in terms of volume) they tapped a large new audience for used books. Originally all they did was take “want” requests and then tried to find copies by using one of the dedicated used book sites. But they quickly realized that this was a large and relatively untapped market and they decided to aggressively go after this business themselves. They, along with Ebay/, have turned used bookselling into a commodity business.

Or at least a part of used bookselling.

Used Books as Commodities

Before the Internet it was difficult for a customer to find most out of print books. It wasn’t because there weren’t a lot of them out there, it was because these books were “hidden” on the shelves of thousands of used bookstores. If it wasn’t available in the handful of stores close to you then it might as well not exist. There were a lot of books, but limited accessibility.

Online bookselling changed that. With places like Bibliofind and Abebooks a customer could “search the shelves” of bookstores all across the country (and the world.) Availability opened up. But the customer base was still relatively small and primarily focused on the collector who was used to searching for books. When started to promote used book sales along with their new books they brought a lot of new customers to the used book market. Now you had a lot of common used books, ready accessibility online, and a large customer base who wants them. Used bookselling became a commodity market. The problem is that this only applied to a part of the used book market; the one traditionally served by the paperback book exchanges and booksellers of that kind. Readers on a budget looking for non-collectible copies of common titles. These customers are much more interested in the content and price of a book than they are of some other criteria commonly looked for by collectors, such as edition and condition.

So we have a large market and a lot of books to sell them. Now the challenge is how to get these books listed online in the easiest (and cheapest) way. It takes time (and therefore money) to create a listing for a book. Although there are a lot of these common titles out there, they aren’t coming from one or two sources; there are hundreds of booksellers each listing one or two copies. Commodity selling requires low overhead costs; the profit on any one copy of a commodity book is small. The profit is in selling lots of them. But since any one dealer has, at best, a handful of copies, the overhead for each bookseller to create their own online listing is high. How can this be reduced?

The Good

Okay, you saw this coming; the answer that has been proposed is ISBN lookup. Most of these common titles are relatively recent releases and have ISBN numbers assigned to them. Bowker (and their International equivalents) have databases of these numbers with information about these books that was provided by the publisher when the numbers were issued. The ISBN number for a book can be used to pull up a lot of details about it; this information can be used to populate an online database. The overhead for the bookseller decreases dramatically.

The Bad

Unfortunately this approach is not without its drawbacks. There are some additional criteria for used books that aren’t found in the ISBN database:


  • Condition – used books can be anything from “as new/unread” to “well worn.” Even for common, non-collectable books some gross indication of condition is necessary.
  • First Printing – there is a large collector market for First Edition, First Printing copies of modern books. Although the ISBN can be used to indicate the edition (all new editions require a different ISBN number) it does not differentiate between printings.
  • Multiple Editions – in addition to being reprinted, common titles often have multiple editions (and therefore multiple ISBNs.) Sometimes these are indicative of different formats (hardcover, trade paperback, mass market paperback) but sometimes there can be multiple editions in the same format. When a particular title has multiple editions (and therefore multiple ISBNs) they can often be easily confused for each other.

Some of these issues can be easily addressed. When dealing with common books a simple statement of condition (Fine, Very Good, Good) is probably sufficient. First Editions can be similarly identified (understanding that with a collectable book you also need to go into much more detail as to the condition.) Multiple editions can be more challenging, although the format (hardcover, mass market paperback, etc.) will often be a differentiator between different ISBNs for the same title. The good news about these kinds of problems is that they can be addressed by the listing services or by the booksellers themselves. Proper use of the ISBN is something the bookselling marketplace can control.

The Ugly

But there are some real nasty problems; those created by the listing services and/or booksellers in assigning ISBNs to a listing and those that exist within the ISBN system itself.

Problems with assigning ISBNs are mostly a result of trying to link an existing entry for a book with an ISBN number. As previously noted many common titles have more than one ISBN associated with them. Booksellers have spent a lot of time creating their book entries and are generally not willing to throw all that work away and start fresh using ISBNs lookups. Since the ISBN is often not noted in older listings, attempts have been made to use the title, author and/or format to assign a number to them. Unfortunately, these attempts often fail, with the result of incorrect ISBNs being assigned to a listing. This in turn angers customers when the book that they receive doesn’t match the online description.

This problem is aggravated when the information in the ISBN database is itself faulty. Back in the early days of ISBNs there were a lot of opportunities for error. I can use my own experience here as an example.

Back in the late 1970’s while still in college I started a small press, The Burning Bush Press, to publish some fiction by a friend of mine, Mark E. Rogers. Although the editions were small and the publisher was a poor undergraduate student, I made an attempt to do it “right.” That meant registering the copyright with the Library of Congress and getting an ISBN from Bowker. (I’ll admit that the only reason I bothered with the ISBN was that I wanted to get our books listed in Books in Print; I didn’t get around to doing this until my third and final title.) I filled out the appropriate paperwork and was assigned a publisher’s identifier and given forms to fill out in order to submit my titles.

Assignment of the title identifier was left entirely up to me. And adding this title to the growing list of assigned ISBNs was a manual process; I filled out and mailed the form to Bowker, and someone on their end had to enter the data into their “system.” The first (and only) book I submitted was a short play-parody, Waiting for Gomot by Officer Joe Beckett (as transcribed from the original crayon by Mark E. Rogers,) ISBN: 0-937528-01-3. Essentially it was the Three Stooges doing Waiting for Godot. A silly piece that had only been performed once, on May 16th, 1976 at the University of Delaware with an accompanying electronic music score by John J. Adams.

You can probably guess what happened. Bowker managed to screw up the listing, showing it as Waiting for Godot. For several years I received orders from teachers wanting copies for a class (at $2 ours were cheap.) Add to this another little problem; if you looked up “Burning Bush” in the Publisher index I was the only one listing. However, there was a Burning Bush imprint used by The United Synagogue Book Services. As a result I would get an occasional order for books about Jewish Dietary Laws.

These are the kinds of things that happened when you had a manual system of entering data. Add to this the potential problems created when the assignment of the title identifier was left solely to the publisher (if I hadn’t submitted the paperwork Bowker wouldn’t know that a title identifier had been used). These allowed for both errors on the transcription side by Bowker and errors on the assignment side by the publisher. Also, as the number of titles grew the larger publishers began to run out of numbers. This meant they had to get a new publisher identifier; unfortunately some chose to reuse numbers instead. This problem was magnified by all of the mergers and takeovers that occurred in the publishing industry over the last 25 years; often the publishers themselves lost track of what ISBNs they had already used.

Today these problems aren’t quite so severe. Computerized databases and submissions have helped, and Bowker now issues title numbers to publishers in groups of five. This makes it easier for them (and the publisher) to track which numbers have been used, helping to prevent the issuance of duplicate numbers. This also helps to minimize transcription errors, although good copyediting on the part of the publisher is still required when submitting information. But the damage has already been done.

The Situation Today

There is a need for a simple and straightforward way of entering information about common, commodity books for online listing. One more focused on the content of a book than with details about a specific copy. The ISBN is a likely source for such information; however as had been noted above, this approach has a lot of pitfalls today. Changes are needed before the ISBN can be used successfully to list and sell commodity books.

And these are coming. New standards are emerging in the new book world that will help. In a future article I’ll talk about these new standards, what they’ll mean to online book community, and some proposals as to how they can be successfully implemented by the online book community.

By: Erwin H. Bush

The Standard: The Journal of the Independent Online Booksellers Association

Check out the Independent Online Booksellers Association Website