Moving Libraries to a Web Services Environment – Issues To Consider
by Todd Carpenter (Managing Director, NISO, One North Charles Street, Suite 1905, Baltimore, MD 21201; Phone: 301-654-2512; Fax: 410-685-5278) <email@example.com> www.niso.org
In April, OCLC released the first iteration of a Web-based service for library management systems. This is the first salvo in what will likely become a radical transformation on how libraries manage their resources — both in print and digital forms — as well as their services. Much like many industries that are moving to hosted or “cloud”-based solutions, libraries are assessing the practicality of running their own complicated back-end office systems, their integrated Web-based user applications, all their discovery tools and the ever growing multitude of information management environments.
What is a Web Services Environment?
In this environment, an organization uses a third party service and their networked information resources to provide information technology, software and services, rather than owning and running all the services in-house. Industry has been moving in this direction for some time, generally referring to such vendors as application services providers (ASPs). A simple example is a Web-based document creation tool such as Google Docs that is used to replace desktop word processing systems.
One service that is frequently cited as an example of cloud-based services is salesforce.com. Organizations that rely heavily on sales teams, who are frequently on the road, need centralized contact and customer relation management (CRM) software that is accessible from anywhere the sales rep happens to be. They have been turning to this service to provide it since it was launched in 1999. Lest one think that Web-based applications are a niche market in software, salesforce.com saw its 2008 revenues top $1 billion. Beyond sales management, other popular management systems in a Web environment are accounting— NISO, for example uses QuickBooks Online — Gmail to replace enterprise email systems, Skype or Vonage for telephony, or even Amazon’s Elastic Compute Cloud that provides processing capacity.
The benefits of using a remote, Web-based platform for information services can be tremendous. The company no longer has to purchase and manage costly servers and networking technologies or address the significant technical issues with controlling access or security, and applying the frequent and necessary software updates and hardware upgrades. Training costs for IT staff to stay current in an ever-changing field can be reduced or eliminated. New capabilities may be available faster as the customer base and competition can drive the supplier to implement new capabilities sooner than an organization might do so in-house.
OCLC’s Plans for a Web Environment Library Service Structure
For many years, people have seen the potential of applying the principles of Web computing to library management systems. Andrew Pace, formerly at North Carolina State University and now Executive Director of Networked Library Services at OCLC, was one of the earliest and most compelling of these visionaries. In 2004, Andrew wrote an article in Library Journal titled, “Dismantling Integrated Library,” (http://www.libraryjournal.com/article/CA374953.html) where he envisioned a structure of interoperable components operating in a Web-based environment. Fast forward five years and Andrew is leading a project to launch the first services of exactly this type of interoperable Web-based library management system.
Drawing from Andrew’s recent presentation on the topic during a NISO Webinar (http://www.niso.org/news/events/2009/interop09/interop09_web.pdf) and from the April 2009 OCLC release on their new strategy to move library management services to Web scale (http://www.oclc.org/us/en/news/releases/200927.htm), OCLC’s cooperative library management system is an extension of WorldCat local and the FirstSearch service. Their release states that the system provides “libraries a locally branded catalog interface and simple search box that presents localized search results for print and electronic content along with the ability to search the entire WorldCat database and other resources via the Web.” What is interesting is the combining of services with integrated holdings and search functionality in a Web-based environment. In addition to reducing the costs of operating these systems locally, the data can be combined with other organizations to further enhance end user services.
Issues to Consider Before Moving Your Services to the Web
It will be important for library managers to consider carefully a number of issues before proceeding down the “cloud” computing path. While the savings might be significant, turning over an organization’s information services to a third party can be fraught with risk. Will the service company provide the same level of service your organization is accustomed to? Management can dictate to its own staff and can dedicate resources to fixing, upgrading or enhancing an in-house system. However, once services are outsourced, there is a range of limitations that the organization needs to deal with. All of these issues can and are rightly dealt with in a service level agreement with the vendor.
Some of the most critical issues surround the data that is now stored on someone else’s computers. Obviously, an organization might not want the actual data to be shared or mingled with that of other organizations or competitors. There are certainly privacy issues surrounding data stored on third party systems, but there are contractual and technology solutions to address these concerns.
Ownership of data is another question. While obviously rights to one’s own data generally (though not always) is a given, who owns the data about the data, such as usage logs and transaction activity? Data aggregation can be a very powerful tool, even if anonymized. The meta-analysis possible when reviewing information across numerous institutions could prove extremely valuable to other organizations, or simply to the vendor itself. One need only look to the MESUR project underway at Los Alamos (www.mesur.org) and some of that group’s work on click streams and usage patterns to get a sense of the power (and financial opportunities) of large-scale meta analysis from crunching data in usage logs. What limitations (or lack thereof) are there on the uses the supplier can make with the data that is created from using its services.
This issue came to a head earlier this year with the release of OCLC’s new Proposed OCLC Policy for Use and Transfer of WorldCat Records http://www.oclc.org/us/en/worldcat/catalog/policy/policy.htm). There was a significant outcry from many in the community about these proposed changes, including from ICOLC (http://www.library.yale.edu/consortia/statement-oclcrecorduse.htm), ARL (http://www.arl.org/news/pr/oclc-policy-20feb09.shtml) and others (http://dewey.library.nd.edu/mailing-lists/ngc4lib/). OCLC was forced by the community to withdraw the initial proposed terms and engage a Review Board on Principles of Shared Data Creation and Stewardship (http://www.oclc.org/us/en/worldcat/catalog/policy/board/default.htm). There is a wide range of applications for which OCLC would like to use the data that it has received from the library community and there is a need to ensure that they have the rights to do so. However, the library community also has a desire to take advantage of the data that they supply to OCLC and that of others, where appropriate. The library community and OCLC need to come to a common understanding about what is allowed and what is prohibited on both sides of the agreement.
On a more distant timescale, there are also lock-in concerns that are slightly more challenging than in a situation where the organization internally manages it’s solution. Software migrations are significant enough when one is dealing with an in-house system acquired from a vendor. However, moving from one Web-based service supplier to another might be significantly more challenging (and costly). Without access to the back-end of the system, customers would be forced into relying on the interfaces and conversion capabilities that a vendor supplies. It is likely that not all of the data especially system-related metadata) might not be extractable in any usable format.
Many of these issues can be addressed in service level agreements, but they need to be carefully developed and attached to any contract for services. Librarians who have mastered the request for proposal and negotiation of license contracts for content now have an entirely new and complex area to learn about.
Why Should Publishers Care About This Trend?
Publishers and other content providers would do well to pay attention to these developments in Web services computing for library systems. At the very least, providing information that is compatible, interoperable and accessible by these next generation library management environments will be an important component of making publishers’ books and journals available to end users. Without easy integration into a library’s workflows, it is far less likely that content will be widely used. Certainly content is king and critical titles will be acquired when there is demand, especially from faculty. However, without integration into the library order processing and management systems and exposure through the discovery and delivery systems, content acquisition and usage could be in jeopardy.
Support of standards related to the exchange of data between publishers and library systems, such as SUSHI, COUNTER, CORE and ONIX-PL, will become even more critical. Their adoption by publishers will become increasingly important as tools to interoperate with and populate information in these new library management environments. Content providers who are already adopting such standards will be better positioned as the library Web services trend grows.