Tor bookshelfOne of the first things I learned upon getting a job in publishing was that (and I quote) ‘Metadata is King’. Now, I didn’t have a clue even what metadata was, but obviously whatever it was, it was important. But why? Below, I’m going to explain a bit more about metadata and why it matters to us as the publisher, and also to you as a reader.

So what is metadata?

Well, it’s all the information about a book that, as a publisher, we send out into the world and that you probably don’t normally think about but actually makes a huge difference.  Primarily there’s the obvious information – Title, Author, Price, Release Date and Blurb – but on top of that we supply keywords, territory information, ISBN, format information and category codes (explained below!1).  If you’ve ever wondered why, for example, an ebook is listed with page length when it doesn’t have pages, or why sometimes, when you’re searching for the new book by your favourite author, totally different books show up … it’s all because of the metadata.

The principal thing to remember about metadata is that as long as everything is correct, you shouldn’t notice it. When it goes wrong, just remember, the publisher probably already knows and is chasing down the person responsible … 

What good metadata means: 

Good metadata means that when you want to find a book, you can find it. It means that all the information we have about a book is sent out to retailers, which lets them list the book correctly, which allows our readers to find that book (see the London Falling book cover and metadata examples here). It means you can type an author’s name into a search engine and their books show up correctly.

In the UK, the overlord (in a nice way!) of good metadata is BIC (Book Industry Communication) who are the book industry’s independent supply chain organisation and who create and maintain standards for the supply of production information.

What happens when good metadata goes bad?

metadata1Nobody ever means to send out bad metadata. Sending out bad metadata is like going into a bookshop and taking a Science Fiction book off the shelf with all the other SF books and putting it with the recipe books.  Nobody is going to look for the latest space opera in a section filled with books about baking cakes.

When metadata goes wrong, it can get pretty ugly. Put a 1 or a 0 in the wrong box, and your book can disappear completely (yes, some metadata is binary, it’s cool like that). And it’s not just a lack of information. Too much information can cause some systems to overload. Have you ever seen a book listing that goes something like this?

AIR WAR: BOOK EIGHT (Shadows of the Apt, Book 8) (Shadows of the Apt 8)

Not only is the formatting all wrong – capslock, there should be a ‘The’ at the start -- but there’s far too much information. This is bad (and, I would like to reassure the author, I made up this example and in reality the metadata for The Air War is a beautiful thing).

What was metadata like twenty years ago?

Honest answer? I don’t really know. Twenty years ago I was too busy watching The X-Files (ok, nineteen years ago if you want to be specific) and listening to Take That to pay attention, but I imagine it was slightly different to today.

Research tells me that in the dark, pre-internet days, there was a traditional split between the metadata for trade publishing (what we do) and library and scholarly publishing. Metadata was mostly limited to the basic information you needed in order to know in order to sell that book: is there a copy of a book in the building? Where is it? And how much does it cost? Because bookselling was based in a much more physical world, it wasn’t until the rise of online retailing that publishers had to start thinking about that dreaded word ‘discoverability’.

At which point, publishers all had to start thinking about information that they’d never had to send out into the world before. Naturally it wasn’t the wonderful world of clean metadata that we have today (on the whole), it took a while to get the hang of it but eventually we got the hang of it. Which leads us to …

How about now?

In reality, good metadata is a really well filled-in record (whether a database or spreadsheet) which leads to a book appearing correctly on every website that lists it (not particularly exciting, I know). But in order to get to that really good spreadsheet, you have to start thinking about a book’s metadata from the very beginning: What category does it belong to? What does the blurb need to say? Is it part of a series? How much will it cost? Where will the book be sold?

Getting the metadata right can make a huge difference to the simple task of getting a book to readers, particularly those who are simply browsing rather than looking for something specific.

What’s the most exciting thing about metadata?

Imagine a bookshop. The doorway to this shop takes you through to a parallel dimension, allowing this bookshop to contain every book from every publisher that’s currently available. It’s an amazing place. The shelves stretch for miles and you could wander for hours looking for the book that’s just right. But where do you start? Hard to say, there doesn’t seem to be an order to the system.

But wait! Conveniently there’s a computer here that lets you search for the type of book you’re interested in. Put in a few words and it’ll give you suggestions. Are you looking for a new release? Do you want something published in the year you were born? Something that inspired your favourite author to write your favourite novel?

That’s the most exciting thing about metadata. It helps us, the publisher, get a book to the right reader.

*  *  *  *  *

1In the UK, Science Fiction is FL and Fantasy is FM, both of which break down into further sub-listings such as Space opera (FLS) or Fantasy romance (FMR). And then, naturally, the US has a different system according to which Science Fiction is FIC028000 and again breaks down even further. Visit BIC and BISAC if you want to find out more.