[Sponsored Ad]
       Advanced Search
 search by genre:
  
display only ebooks  audiobooks display only ebooks  audiobooks  
  search within genre:


newsletter archives

August 2012 - Interview with Laura Dawson: Integrating Metadata into the Publishing Process

Reprinted from original interview by Quill & Quire.

What is good metadata?

The concept of "good metadata" is going to vary depending on who's using it. What a distributor expects to see (and needs to see) is going to be different from what a consumer expects (and needs) to see. It's critical to keep in mind that metadata is actually a form of communication, and you need to think of your audience. Good metadata is what gets the appropriate message across clearly and without ambiguity; it's a publisher's announcement that the book exists; it's the accompanying details to further describe the title for whomever needs to move the book along the supply chain. Not everybody in that supply chain is going to make the same use of the same data points.

Why is good metadata important for publishers?

Metadata is often the first indication to the world that the book is being published, so getting early word out is dependent upon communicating clearly. As more shoppers move to online purchasing, metadata is the only way they will be able to find the book they want. In the case of ebooks, because there is no physical object, metadata is literally the only way a consumer is going to know that the book even exists. So for ebooks, it's not just important - it's critical.

What are some of the most common barriers publishers face (or missteps they make) in putting together good metadata for a title?

One of the biggest missteps is not putting themselves in the shoes of the folks on the other end who are receiving the data. Good metadata is the result of relationships between the publisher and the rest of the supply chain. Understanding your trading partners' needs is crucial. Another common mistake is treating it as "someone else's job." The truth is, everyone in the publishing process has something to say about the book. Assigning metadata entry to a summer intern means that a lot of that knowledge will be lost. If the entire publishing house is actually using the metadata in order to put out the book - if everyone has a hand in it and sees it every day - it's likely to be pretty solid. But if you shunt it off into a separate function - the "metadata manager," for example - that further silos the process and virtually ensures that there are going to be problems.

If everyone involved in the title should have a hand in contributing to its metadata, how should this information get stored?

Ideally in a central database that each division has access to. People learn a lot from one another, and allowing them to view as much information as possible means that people will notice errors and correct them. That said, you obviously don't want designers messing around with financial statements, so within that central database there have to be certain levels of permissions.



**********

Did you like this article? Let others know! Add it to:

Digg.com |  Reddit.com |  del.icio.us |  StumbleUpon.com

**********

Back to the Archives



HOME    ABOUT US    FAQ    ADD BOOKS    ADVERTISE    NEWSLETTER    BOOK VIDEOS    YOUR VIEWS    LOGIN    CONTACT US

[Sponsored Ads]
Place Your Ad Here








Tell us about yourself

My bookhitch Experience