BookNet strongly encourages all publishers to be the creators of their metadata. You control it and supply it to those who need it. This post is trying to address the problem a publisher may have when when their new distributor doesn’t support metadata distribution and they as a publisher have become responsible for representing their distribution to Canadian retailers. That said, most of it applies to any publisher and its concepts would also apply to retailers.
Let’s start at the most general
Published material is not distributed in isolation. Anything published is established in a final form that, while it might be modified at a later date, becomes locked — published — at a specific point. It’s a thing. A published book is the creation of a publisher who has crafted content to be its best representation. That book is duplicated and distributed, though the exact means depends on its format and the options available to the publisher.
Book metadata describes that published book and can include: the exact title and author as appears on that book, its size and weight, whether it has shiny endpapers or a spine that forms an image when displayed spine out with other books in its series. That’s before you open the book. A whole lot more needs to be said about the contents before you even get to publisher’s business terms and distribution arrangements. Book metadata, information about a product, can also be said to be published. It describes the book as best it can be at given time but even after the “thing” becomes a published book the metadata still changes. There will be multiple small changes as the book or its author wins accolades, as the publisher develops new markets to sell in, releases alternate products with the same or related content, updates the price and provides promotional opportunities to consumers, etc. Even if it hasn’t been updated, a book record includes time-sensitive status information. If you have a two-year-old book record that says the book is available: Is it? Is it, really? It would be a lot more believable if that record had been updated within the last six months.
Book metadata should provide a potential reader everything they need to know to make a buying decision and it should support the business needs of retailers. It's time sensitive and designed to support other businesses who support a publisher’s sales. There are a huge number of books published each year. Your metadata has the potential to be seen by far more people than can touch one of your lovingly crafted books. The ideal is that it appears in a retailer’s lovingly curated search results built from the exceptional and regularly updated metadata provided by a caring publisher.
Book metadata is published but it exists to support the business needs of a supply chain (best understood as a market) where it’s sold. It's traded within a system of standards that are, in a sense, arbitrary conventions established and agreed to in order to facilitate sharing information between businesses. Every business tracks its needs in a way that enables it to meet its goals, but it “maps” (converts might be an alternate term) its internal knowledge into common terms that other companies can read and act on. Metadata that accurately matches standards decrease costs and increases sales across all trading partners including for the publisher.
An ISBN is a unique internationally assigned identifier that represents a single saleable product in a specific format.
An ONIX record describes that same salable product
What that means is our “thing” — the published book — is going to be described in how you sell it: via an ISBN which is like an order number. Using an ISBN means the ordered product is always understood as specific format and specific content as published at a given time. It always gives you back the same salable “thing” when the book is available as paperback, hardcover, digital EPUB, digital audio. Each one has its own ISBN and each one has its own ONIX record.
An ISBN-based record is the source for what a retailer (or any company) displays online. Online display is typically related to the sale of a specific format, though increasingly retailers group formats and toggle the format specific details within a book specific page. Properly done ISBN-based metadata can support both book and format.
ONIX for Books is a metadata standard where its organizer, EDItEUR, in consultation with book supply chain groups across multiple nations, has made a created set of communally sourced rules that allow “standardized” communication about books. It starts with the ISBN and layers on a system of code lists and values that allows something like 200 concepts and maybe 4,000 unique data points to be traded. (Most book records use a much smaller subset that that!)
Book metadata also includes ancillary materials, normally files named by ISBN, and at a minimum should include a cover image for every ISBN you sell. Other types of ancillary material varies a lot by specific type of book and readership but can include interior page images (as an image file or as PDF sample excerpt), tables of contents (can also be embedded in an ONIX file as formatted text), and other types of material that can help a retailer sell to their clients: author photos, specialized sales sheets and TikTok or YouTube videos and so on.
ONIX for Books supports download URLs that can link to these materials.
Book metadata can be traded in other standards (things like Dublin Core exist) but EDItEUR’s ONIX for Books standard is by far the most complete book metadata standard, and it’s one designed to be a container for other book standards. So ONIX can “contain” BISAC and Thema subjects, include a GTIN-12 (a.k.a UPC) or an ISNI identifier, or information specific to Japan or the EU. EDItEUR typically adds support to ONIX prior to your need to trade it — and if you have a need to trade information that isn’t supported talk to BookNet Canada about it. We organize ONIX for the English language supply chain in Canada. If it’s needed and your trading partners agree it can become part of the standard.
ONIX is designed to be distributed in XML-based files as “feeds” but as a metadata standard its definitions are typically used in all kinds of book related communication. You can trade book metadata in Excel but it's most likely using concepts and codes as defined by EDItEUR. The important thing is that a published book has a lifetime that ends when its publsher no longer supports its sales. That means the metadata describing a book cannot be static or a one-time event — at least if the business need of the companies selling the product are considered.
“Feeds” are what we call regularly submitted metadata that follow the changes required for a retailer to be up-to-date in maintaining their ability to sell a book.
For the purposes of this description we going to use “ONIX feeds” to refer to the specific XML-based ONIX file that is what most business use, but any feed you have been asked to support is likely still referencing “ONIX” as a metadata standard.
Distributing book metadata
Now we have a place to try to answer just what it means to distribute book metadata and what your responsibilities might be.
It has to start at the obvious: The distribution of metadata is dependent on the contracts a publisher signs, as well as the contracts held by those companies. Everything will vary by those contracts and the associated markets served. The obvious leads into our most generic piece of advice:
Regardless of who or how your book metadata is published it’s the publisher’s responsibility to understand who’s getting their book metadata and why. If you’re relying on a trading partner to distribute your data you should have some idea of where and any limits that distribution has.
This is why BookNet strongly encourage all publishers to be the creators of their metadata. It provides you the ability to both control it and supply it to those who need it as necessary.
Anyone who needs it should get it should get a metadata feed.
That would include all retailers who sell your books and were supplied by your previous distributor, your current distributor, or by you directly.
It should include all the terms needed to sell your books, while that can vary depending on the markets you and your distributor sell into (libraries, schools, and so on may need consideration and pricing) you should should make sure that you are using the correct code value(s) to support your distributor’s discounts:
Ask for their discount code(s — it may vary by categories) to appear in ONIX’s DiscountCoded composite for each price supported.
If cross border shipments will occur, you may to support data points like Country of Manufacture and various Product Classification Type (List 9) codes for tariffs and harmonized commodity coding. It might even require a Country of Publication code. Your feed will need to meet their needs.
You should be prepared to add new recipients to your metadata distribution list. We would recommend that a contact name be provided on your website and that you and your distributor regularly compare notes.
You should be able to support “feeds” that update retailers on changes — new books coming in, old books no longer supported, reprint and publication dates maintained so any retailer promises made to their customers can be met. A “feed”:
can include manually supporting the online entry forms of a major retailer;
can include creating a spreadsheet to support a small specialty store; and
always includes distribution of ancillary materials — cover images at least in addition to metadata.
Your goal should be to use ONIX to organize your metadata to simplify the former while working towards supporting all of your accounts through a direct and simpler ONIX file feed.
If your new distributor has contracts with major retailers then you should be able to supply ONIX feeds to them, but those feeds may need special considerations. (Ideally all companies can and will accept the same ONIX file in their feeds, and it should be possible in most cases, but there may be exceptions.)
Small retailers can increasingly work with book metadata suppliers and services that enable them to those systems to absorb metadata. Look to support suppliers of services.
Do not forget that third party data aggregators exist and the book supply chain includes a wide variety of business such as libraries, wholesalers that support sales to libraries, universities, and schools. Expanding your business and its readership, pretty much by definition, will result in expanding who gets your book metadata.
Foreign distribution needs to be considered
How will your metadata feed interact with other feeds in other markets that are also representing your ISBNs?
Do your ONIX feeds need to be supplied to different distributors in foreign markets?
Note that if you send CAD currency information outside of Canada that it's a commonly used marker by US distributors to determine where they should send their metadata. They may not represent Canada, their data may correctly (or incorrectly in that it may not follow best practices for markets in ONIX metadata) identify they don’t supply data to Canada, but by sending them a CAD price they will send your book metadata to Canadian retailers and let them sort whose data is right. Needless to say, “Results may vary.” But it’s easily solved by simply not supplying CAD currency to a foreign business not responsible for supplying your books to Canada. It’s just a good practice to send data that’s needed and leave off anything not needed in that market.
How to communicate distribution changes
Anytime a publisher changes distributors they should announce it to their trading partners and include a clear statement of the date the change happens and when and who is accepting returns with any cutoff date.
The previous distributor (or the publisher acting for their distributor) is responsible for sending a metadata feed on or just before the change date with a Product Availability List 65 status for Not Our Product. This will "close" their records for any trading partner doing business with them. There's no reason to repeat this as it will just interfere with the new distributor, but it's normal to supply newly NOP and OP records in "full file" feeds for six months or so after the change.
The new distributor (or the publisher acting for their distributor) is responsible for sending a metadata feed on or just after the change date with a Product Availability List 65 status for as Active sold (or forthcoming). This will "open" their records for any trading partner doing business with them. In order to make sure that everyone's records get updated it's a good idea to repeat that feed in a month. We have a whole blog post about this if you want to learn more about it.
Where to begin
Outside of making sure you’re covering major retailers your distributor works with, be sure to supply data to:
Bookmanager: Their POS software and metadata distribution service supports many independent bookstores in Canada and the US. If you have split Canada and US distribution, make sure that they are receiving the data they need from the appropriate source.
To make sure that BookNet Canada products like CataList and SalesData have your data, be sure to send your metadata feeds to:
Bowker / Clarivate: Bowker’s database is one of the few comprehensive ones in North America and they support the Canadian market distinct from other markets (including the US). They will almost certainly be receiving a feed from your US distributor if that’s separate so they'll need to know that your feed represents Canada. Bowker’s metadata currently underlies SalesData.
BiblioShare: BookNet Canada’s data aggregation service wants and needs feeds for any ISBN sold in Canada, and absolutely should have a feed from any Canadian publisher. Our APIs support Canadian book wholesalers, libraries, and special marketing services like 49th Shelf. Our products include Bibli-O-Matic, a browser extension that provides fast look up for the Canadian book industry, and we support a Shopify account plug-in that pulls book metadata into their stores. And of course, our data underlies CataList, our well loved and used e-catalogue service.
Questions? Contact referrals? Did we miss something here? Get in touch!


We’re seeking book industry professionals who are actively experimenting with AI tools in thoughtful and ethical ways.