INDUSTRY STANDARDS

Use Better Data to Boost Your Sales

by Michael Healy

The message is simple and uncontroversial. Good product information helps everyone sell more books. But according to leading booksellers, many publishers—large and small—still have problems maintaining good-quality book data. Determined to help, the Book Industry Study Group has introduced a service that measures data quality and provides feedback to help publishers meet industry requirements.

The new service, launched in February, is called the Product Data Certification Program. Essentially, it allows publishers to send their data files to the BISG office, have them examined, and get information about their quality. The service is open to all publishers—it’s not necessary to be a member of BISG. It is free in this first year. And a niche publisher—Waterford Press, which produces reference guides that introduce novices to travel, languages, science, and nature—was one of the first three companies to win certification.

How It Works

Publishers wishing to be part of the program complete a short application form and deposit a data file on BISG’s FTP site. You need to submit either a properly formatted ONIX XML file or an Excel spreadsheet that is compliant with a template available from the BISG office. No other file types will be accepted. Your file must contain data records for all titles you published in the previous 180 days and all titles you will publish in the next 180 days. No titles outside these parameters are analyzed. (See “Help Is at Hand” below if this sounds like a tall order.)

Once a data file in the correct format has been received, its contents are analyzed for the presence or absence of 17 mandatory data points for prepublication data and 30 mandatory data points for postpublication data. These data points are described in the program’s documentation and are the essential parts of any good-quality product data record as defined in BISG’s Product Metadata Best Practices.

If 80 percent of the records in the file contain the mandatory data points, the publisher is eligible to be formally certified by BISG and will receive a detailed scorecard showing its performance.

But the analysis and scorecard make up only half the program. After the file has been analyzed in terms of the key data points, it is sent for review to a panel of data specialists from Baker & Taylor, Barnes & Noble, Bowker, Ingram, and the Library of Congress. The reviewers do spot checks on the contents of the files, check the validity of the XML if appropriate, and provide aggregated feedback, which we will send you along with the scorecard.

Under the terms of the program, publishers are certified for one year and are required to resubmit data files annually to maintain their certified status. We decided not to charge publishers in the first year in order to encourage adoption, but in subsequent years there will be a small charge to cover BISG’s administrative costs in running the program.

Publishers interested in being part of the program or learning more about it should visit bisg.org/documents/certification_productdata.html, where you will find a program description, FAQs, and details of the application process.

Is This Really Necessary?

Is there a single publisher who hasn’t heard or understood the message that good data spurs sales? It seems unlikely. After all, for more than 10 years, organizations like BISG have not only been preaching that message at every opportunity, they have been giving publishers the information they need to understand what good product data is and tools with which to create and maintain it.

Yet after all this time and after all the industry education, standards development, and best practice guidelines, many booksellers report that the overall quality of product data in our supply chain remains stubbornly low. Joe Gonnella of Barnes & Noble, one of the most determined and enthusiastic advocates of higher standards, is among those booksellers. “While some individual publishers have made progress on the quality of their data for new titles,” he says, “overall improvement of new title data quality in the past three years has been marginal.”

Everyone who wants to sell books should be concerned by this conclusion, especially because it is a judgment supported by extensive evidence from Barnes & Noble’s bookstore systems and its Web site.

If the booksellers are right that the quality of product data is not improving quickly enough—and there is plenty to suggest they are—why do we, as an industry, have such problems? It’s partly to do with the growth of new publishers and the explosion of available titles. Figures obtained from Bowker indicate that there are more than 120,000 active publishers in the United States and more than 3.3 million available titles in the supply chain. Figures in both categories are growing fast year-on-year.

Barnes & Noble can verify this growth; recently it reported that it has more than 7.4 million active or inactive titles in its database, an increase of 33 percent in three years. New titles are a factor—publishers released 800 new titles every day in 2006—but so is the industry’s growing backlist. With developments in short-run printing and print-on-demand, more and more titles are feeding the long-tail phenomenon so widely discussed in the industry instead of going out of print.

Creating and maintaining high-quality product data for this volume of titles is clearly challenging for publishers.

Help Is at Hand

Perceptions about complexity are also part of the problem. Some publishers find it difficult to understand what good product data is and how to go about creating it.

It’s easy enough to understand the three attributes of good data—accuracy, comprehensiveness, and timeliness—but there can be no doubt that some publishers struggle to achieve them consistently, especially for large numbers of titles.

Still, the basics are not difficult and available tools—including not only the new Product Data Certification Program but also BISG’s Product Metadata Best Practices—can help a lot.

The inquiring or despairing publisher’s first response should be to consult Product Metadata Best Practices, which can be downloaded without charge from bisg.org/docs/Best_Practices_Document.pdf.

At 86 pages, it’s not a quick read, but it clearly sets out 31 “core metadata elements” (aka data points) that make up a good book data record, and it explains how to express each of them properly.

Understanding the various parts that make up a good book data record is the essential first step toward your ultimate destination—exchanging accurate and timely data with trading partners.

The second step is working out how to store and send the data. Many publishers in the United States and overseas have adopted the international standard called ONIX for Books to help them with these challenges. ONIX for Books is best understood as a set of tools that enable publishers to express and send rich product data in XML to their trading partners.

BISG manages the U.S. ONIX group through its BISAC Metadata Committee and routinely provides advice to publishers interested in adopting ONIX. Hundreds of publishers of all sizes are now using it to send their product data to retailers and other trading partners. Some are doing it independently, others via one of the intermediaries that help publishers manage their XML data, such as Quality Solutions and NetRead.

This ought to reassure those of you still struggling to organize and distribute good-quality book data that sources of information, advice, and support are available and that they have helped a lot of publishers improve the accuracy and reliability of the data that are so critical to the task of selling books.

Omens of Success

Services like BISG’s Product Data Certification Program (PDCP) that have been introduced in other countries, including Canada and the U.K., have proved to be popular with publishers and have clearly been instrumental in raising the overall quality of data in their supply chains.

Of course, everyone recognizes that it takes real partnership to achieve the best possible product data. It’s not all down to the publisher. Booksellers, wholesalers, and data aggregators play a vital part in getting the best data to customers in a timely way. That’s why BISG will be preparing a set of best practices for data recipients later this year.

By introducing PDCP, BISG and the companies participating in the certification panel are demonstrating their determination to drive up the standard of the book data in our industry, and to do it by offering practical tools and expert advice to publishers who share that determination. Let’s not forget—it’s all about selling more books.

Michael Healy is executive director of the Book Industry Study Group, a not-for-profit research and technology standards organization whose members come from every sector of the publishing community. BISG is a forum for managing change and enabling interaction among publishers, booksellers, librarians, wholesalers, manufacturers, authors, and suppliers. For more information, visit bisg.org.