Implementing the System

Last updated September 27, 2014.

After discussion regarding implementation options, the original four libraries agreed to get started right away. Librarians from the four libraries that had participated in the initial evaluations worked together with the FLO staff to begin the implementation process since we already had some familiarity with the system. The idea was to learn more about the system and train the other libraries once we had a thorough understanding.

Phase One: From Trial to Early Adopters

The initial group knew something about what features existed and had seen the system in action whether as trial participants or evaluators. The group attempted to share one consortium-wide instance, but that idea quickly proved problematic due to the sensitive nature of information, such as logins, costs and variant workflows. At the same time, we also shared some resources that the FLO Office staff administered, and we wanted to eliminate the need for duplicate data entry.  As illustrated below, the current strategy used a hybrid method for recording consortial-wide and library-specific information. The FLO staff maintains one consortial instance of CORAL that is used to feed shared information to the library-specific CORAL instances.   As a group, we continued to use the sandbox to test different options for problematic situations. We could easily compare several methods for accomplishing a single task.

While continuing to populate individual instances, we also created shared, customized documentation, unlike vendor-supplied manuals. Librarians at each site entered local data into the institution’s unique instance and met regularly to review those experiences. We used our community documentation site to share confusing experiences and document areas that benefited from official policy. This online documentation acted both as a means of communication between meetings and as an agenda for those meetings. When several people found different solutions to accomplish the same goal or when one of us came across unique circumstances that challenged us, we had long discussions online, on the phone, or in person to work out preferred solutions. For example, group members agreed upon a method to distinguish between consortially acquired eBooks from one company and eBooks that an individual library independently acquired from the same company. FLO staff members recorded these decisions for easy reference later. The initial trial period also included the development of a common Field Dictionary, and we were able to expand this during subsequent use. Most new elements represented additions to the system, such as the elaboration of the roles an organization could play in the e-resource chain. Other elements clarified terminology. In some cases, individuals were using different terms for the same idea, while in other cases individuals were using the same word to mean different things.   All of this clarity of communication led to common understandings, and best practices.

Phase One implementation was completed in January 2013, even as we continue to build shared documentation, best practices and additional functionality.

Phase Two: Mentoring the Next Libraries

By the end of Phase One, we had several individual instances of CORAL well on the way to being fully populated. We had a consortial instance that included information about organizations and resources shared by all, and we had communal documentation based on common understandings. We also had a group of experienced users ready to become mentors.

Phase Two began in earnest in February 2013 for the remainder of the FLO libraries including Emerson College, Massachusetts College of Art and Design, Massachusetts College of Pharmacy and Health Sciences, Museum of Fine Arts, New England College of Optometry, and New England Conservatory. In the absence of vendor support, this first group trained and served as consultants to the second group of FLO Libraries. The early adopters and the FLO staff members created a series of in-depth training sessions, complete with assignments, to help the second phase participants think through each of the implementation steps. In the initial meeting, early adopters covered reasons CORAL might be useful to the remaining libraries and gave a broad overview of the system to bring everyone to a shared understanding. In subsequent meetings, the first group and FLO staff members created specific topical training sessions on each piece of the system, and provided sandboxes for each institution. Each training session included summaries of the best practices some of us had worked out as well as thoughtful discussions about how those practices might be modified in the future. The early adopters shared recommendations about overall implementation strategies and detailed information about how to use certain features. These helpful hints provided a clear path for each new library.

The early adopters also warned the new participants about system idiosyncrasies, and introduced them to the bug list on the CORAL GitHub website. For example, early adopters explained how adding an item in the licensing module before creating a corresponding entry in the organizations module would create ghost entries and showed them where this was found in the GitHub list. Paid company trainers might not have been so frank about these types of quirks, and vendors often do not supply an easily accessible bug list.

Throughout the training sessions, we also discussed mapping each of our existing workflows to the CORAL system. Follow-up exercises reinforced the in-person training. When new users had questions, the early adopters guided the questioner through potential solutions or helped review the documentation. As hard as this is to imagine, some of these questions were about issues the early adopters had not yet encountered. In those cases, both groups devised new strategies together and incorporated the decisions into the documentation, and our Phase Two adopters became contributors to the growing documentation collection. The FLO office and the early adopters combined practicality with philosophical considerations while developing locally specific training sessions.

At the end of the training, each Phase Two library was given its own live instance complete with consortial-wide date and could begin entering actual data. In addition, FLO staff members eased the implementation process for a few Phase Two participants by transferring content from older SQL-based systems into the appropriate CORAL instances. Phase Two librarians were ready to run with the system.

Member libraries have progressed since then at different rates.   Some libraries are still transitioning from previous electronic resource management strategies to sole reliance on CORAL. Many are using CORAL in conjunction with other tools, and a small number have made it a priority to implement CORAL more fully in the near future.   Although Phase Two of the implementation has officially ended, for many of us, CORAL adoption is a work in progress. As a benefit of participating in Phase One, the early adopters also had the advantage of being able to retrieve e-resource information more quickly since it was already stored in individual CORAL instances. A few libraries are now exclusively using CORAL and have left previous management and tracking methods behind.

Phase Three: Sharing the Wealth

Those of us who were early adopters were truly local experts. We knew the CORAL system as implementers and mentors. A proprietary vendor’s support staff, on the other hand, would have simply applied general knowledge about that system’s overall functionality. A vendor support staff’s knowledge would not have reflected local choices. We also knew the Phase Two participants as colleagues with whom we already shared systems and resources. Through previous experience on a number of consortial committees, librarians in both groups had already established relationships with each other. The early adopters had background knowledge that enabled them to anticipate specific individuals’ concerns and interests. This familiarity made it easy to tailor training sessions for and respond to questions from the Phase Two participants. The early adopters acted as dedicated mentors who could often respond more quickly than vendor support staff, who tend to represent a wide number of customers on an ad-hoc basis.  This internal support was a tremendous advantage to those of us who were Phase Two participants. This mentorship role did not create an unwelcome burden for the early adopters. The additional time was required but had the benefit of strengthening existing relationships.

Because CORAL is open source software with no restriction on the number of instances we can run, this allowed us the freedom to use sandboxes in a variety of ways. Sandboxes were easy to build and did not require significant time of the FLO staff.  The sandbox provided a stress-free environment for the participants to explore the system, experiment and apply new knowledge. 

Using an open source system has been beneficial in allowing consortium members to implement the system on their own schedules, but detrimental in that there is no external vendor pressure to complete training and implementation. Getting an Electronic Resource Management system up and running is a large project, whether the ERM system (ERMS) is proprietary or open source. If FLO had chosen a proprietary ERM system, librarians at each member institution would have faced a similar amount of work. We would still have needed to decide how to make use of such a system. We would have needed to expend the same amount of time and effort to enter the institutions’ data. However, we may have had to do all this on a vendor’s schedule instead of our own schedules. The downside of this is that the lacking sense of urgency may prolong the process.

The relationships among participating librarians that developed during this project remain strong. We now know even more about each other’s responsibilities and workflows.   These relationships facilitate – and are facilitated by – the practical aspect of maintaining CORAL. We have regularly scheduled meetings where consortium-wide concerns are discussed. We supplement those in-person meetings with messages on the internal listserv. Both the in-person meetings and the virtual conversations also allow participants to share new issues and new ways to extend the system. The CORAL implementation project strengthened bonds among FLO members and gave all of us a strong foundation for understanding the larger ERM environment.

We tend to rely on each other first, rather than on the wider CORAL Community for most questions. This inclination comes in part from the strength of the relationships that have organically grown as we worked out shared and local practices. Rather than send a question to the larger CORAL group, we send questions to each other because we know FLO’s customizations. In effect, FLO has developed our own internal community that we use before going to the larger body of CORAL users.

We also realize that we need to participate more in the larger CORAL community. However, while a few individuals in that community actively respond to questions and comments, based on the number of CORAL listserv subscribers, we know that many others don’t respond. CORAL discussions are dispersed across GitHub forums, the listserv, and a few other locations. It takes effort to track all of the conversations, as no single platform is definitive. This lack of a centralized place for interactions also contributes to our habit for internal FLO conversations. 

At the same time, FLO members have extended some support to others outside of the consortium as we continue to learn how best to participate in open source communities.  FLO collaborated with other institutions to transform the locally developed field dictionary into a CORAL glossary. A few librarians volunteered to update the public version on GitHub as necessary. FLO members responded, and continue to respond, to messages on the wider CORAL listserv to offer advice to others who are considering CORAL or who have questions.