The BISG’s Metadata Research Project

Today I listened to the first of three webinars hosted by the BISG about metadata. The series is called Evolving Metadata Practices, and today’s webinar was on A Research-Based Understanding of the Current Metadata Landscape:
Full Results from BISG’s Metadata Research Project. 

Presented by Brian O’Leary, the founder and principal of Magellan Media, the webinar focused on giving an overview of a recent publishing industry survey with 125 respondents. These respondents included publishers, retailers, wholesalers, industry manufacturers, and digital-only people, from both the U.S. and Canada.

Note: Like my recent post, Understanding Metadata, this post refers to metadata that goes to retailers and websites, not metadata within an epub file.

To see other organizations’ findings on metadata, check out “Streamlining book metadata workflow” by NISO and OCLC, and the “Link between metadata and sales” by Nielsen, 2012.

Brian O’Leary said that “There’s enough data points that show that good metadata helps grow the pie.” In other words, metadata definitely helps increase sales.

The study found that many publishers treat a book’s publication as the end date, though this is starting to change.

Here is a list of other findings from the study:

• a complete metadata feed is not the same as an accurate one
• inspection for accurate metadata is limited
• many retailers obtain their metadata from more than one source, which can lead to confusing metadata (especially if the metadata is different in the various sources) and may frustrate consumers (causing a loss in sales)
• most retailers turn around new files in two business days
• there are separate feeds for physical and digital books (which can also lead to confusion)
• in the U.S., ONIX 3.0 is off to a slow start (but there’s good headway in Europe)

Some metadata elements are often modified. Examples include BISAC codes, age range, styled text, weights, measures, and carton quantity (mainly for Baker & Taylor, and Ingram). However, some metadata fields are not used to their full potential, such as BISAC codes and photos. Also, publishers are frequently concerned with the accuracy about the following pieces of metadata: frequent concerns: page count, age range, on sale date v. pub date, territory, territorial rights, and conflicts with contract (different recipients interpret data differently, which may cause problems).

So how can the industry improve metadata? According to Brian, there are a few ways:

1. compare metadata to actual product
2. create stronger feedback loops
3. confirm shared metadata definitions
4. articulate what happens with updates
5. improve transparency on alterations and modifications (this is a “hot button issue” because many data senders have expressed frustration with having to replace data without explanation).

Brian offered a few ways to futureproof metadata:

1. automate data workflows
2. prepare for more frequent updates (especially price)
3. harmonize metadata workflows for print and digital
4. discontinue use of style tags
5. engage new supply-chain entrants

As a basic rule of thumb, good metadata comes out of good business practices. Also, we live in a more dynamic environment, so we need to have more frequent changes to metadata–it’s not a static thing!

For more information on metadata, read The Metadata Handbook, by Renee Register and Thad McIlroy. It’s a guide for creating and distributing metadata and will be available this summer.

Want more? Read Understanding Metadata and ONIX and Metadata.