Amazon Catalog Updates

Did you know that Amazon offers a facility to make corrections to its catalog? Somewhere in the past few months someone mentioned this to me and I tried it out. (Unfortunately, it has been long enough now that I've forgotten who told me; if you are the one, please fess up in this post's comments section. It was Ron Murray from the Library of Congress. Thanks, Ron!) And it works! Is this a model for crowdsourced corrections to library data?

Here is how it looks from a user's perspective.

Step 1. Finding something to correct

Amazon has a pretty good catalog, so for the purposes of demonstrating this feature it took a while to find a record to correct. I used the suggestions from Typo of the Day for Librarians for ideas of errors to look for in the Amazon catalog. One of the suggested typos was Sucess*, etc. (for Success , etc.), and I found a record for How to Talk to Anyone: 62 Little Tricks for Big Success in Relationships in audio CD format with this misspelling. As this image shows, the original title was "How to Talk to Anyone: 62 Little Tricks for Big Sucess in Relationships"

Step 2. Making the Correction

In the "Product Details" section of the Amazon catalog page is a link to "update product info"

Following that link takes you to a form that is prefilled with all of the information from the Amazon catalog. You can make your corrections here and provide citation URLs to reference the source of the correct information. (In the excerpt of the form on this page only the Title and Reference sections are show. Click through the image to see the full version of the form.)

You are given a chance to preview your changes before submitting them. Note in this case that the reference URL I'm using is actually a link to the cover image for this item at Amazon. A bit of neat symmetry there, I figure.

After submitting the changes, you get a nice "thank you" from Amazon for making their service better.

Step 3. Getting Confirmation from Amazon

After a bit -- mere hours in my case -- Amazon will send you a confirmation back that the correction has been accepted.

From: "gfix-noreply@amazon.com"
To: "peter@OhioLINK.edu" Subject: Your Amazon.com Catalog Update Request

==== This is an automated response message - please do not reply ====

Thank you for using the Catalog Update Form to send suggestions for

How to Talk to Anyone: 62 Little Tricks for Big Sucess in Relationships (ASIN 1593160267)

Your update has been accepted and processed. It will appear online within the next two to three business days.
Attribute: Title
Current value:
How to Talk to Anyone: 62 Little Tricks for Big Sucess in Relationships

Your suggestion:
How to Talk to Anyone: 62 Little Tricks for Big Success in Relationships

Data accuracy is highly important to us. We appreciate the time you have taken to submit your updates to us.

Best regards,

Catalog Department
www.amazon.com

And if you go to this product page now you'll see the title has been corrected.

Would this Work for Libraries?

Now Amazon must have some resources backing up this service to do the verification of submissions. And it makes sense for them because corrected metadata makes it easier for their products to be found and purchased. If libraries were to consider providing an equivalent service for our metadata, could we justify the costs? Is this a good use of our time and effort?

If we were to do it, I think it might have to be done by a bibliographic utility like OCLC who has ways to push the updated records to member libraries. Otherwise we run the risk of diluting the corrections across many individual library catalogs. Interestingly, this sort of user-generated correction facility one that the Open Library already provides. (Open Library is a wiki-like service that offers the ability for anyone to make changes to its records, much like how anyone can edit articles on Wikipedia.) So between Amazon and Open Library there is a continuum of workflows of mediated corrections to unmediated corrections for us to consider. This scheme, of course, begs us to consider the notion of distributed version control systems for handling our bibliographic data so that changes can be merged across many sources.

Lots to think about...