D R A F T : This proposal is still a draft under discussion. It may change substantially before it is put into effect. It may even be dropped altogether.

OCLC Research (OR) works with the community to collaboratively identify problems and opportunities, prototype and test solutions, and share findings through publications, presentations and professional interactions. The RLG Partnership, supported by the full capacities of OCLC Research, focuses on making operational processes more efficient by shaping new scholarly services and directly engaging senior managers. OCLC has found that the website it presents for its research efforts at http://www.oclc.org/programsandresearch does not convey clearly enough the story of OCLC Research and the RLG Partnership, the activities of its staff, or the ways the community can join in the efforts afoot.

This spring I worked with OCLC staff to begin a process of rebuilding the OCLC Research web presence. The process included surveying the staff, deconstructing the current site, developing some use cases for the new site, making sure current content and new use cases were accounted for in a new functional inventory for the site, building a category grouping and navigational map for the new site, and finally developing wireframes for a new site. Along the way we affirmed that the new site would be built in the existing enterprise web environment at OCLC, Serena Collage, a system that has been discontinued by its vendor, but will likely continue at OCLC until at least the end of 2010, if not longer.

I have now been asked to help OCLC staff implement the new wireframes on the OCLC corporate Collage CMS. The managers of OCLC Research would like to see the new site ready for public rollout by the end of this summer, so waiting for OCLC to select a post-Collage CMS is not possible. While Collage will make certain tasks simpler, the fact that so much OR content is already in the system will also make migration more difficult. For example, the URL http://www.oclc.org/research is already in use on the site, making repurposing that address for the new site an intricate dance. Migrating existing "activity" and "publication" pages will be particularly difficult. Some on the web redevelopment team have estimated that a migration will take at least six months. I am being asked to help the team accomplish this goal in three months.

I envision three distinct phases to this effort, each roughly one month in duration: planning, execution, and cleanup. In order for us to meet the goal laid out by OR management, every team member and the OR managers themselves will have to push very hard over the summer to be responsive and make quick and firm decisions. Cutting this process from six months or more to three months will require sacrifices that will, in some cases, hurt. The reward will be a much more functional site by the end of summer 2009.

The Planning Phase

While our development process answered many questions, it also raised others. We need to determine the answers to these questions before we can begin to build the new site. I will sit with the team for an initial brainstorming meeting where we will lay out the major questions still facing us as we move forward with an implementation. This will include the drafting of templates for certain standard contributions to the site. We will each take some of these questions away as homework. After two weeks we will reconvene to share answers and make sure vital follow-up questions are addressed. At this second meeting we will also develop a draft timeline for the implementation.

Deliverable: Wiki pages for tracking questions, answers, and templates.

Deliverable: Timeline for implementation with responsible parties for each task.

Timeframe: June

The Execution Phase

The third meeting would transition us into the execution phase of the implementation. We will finalize the timeline and templates, identify responsible parties, and document the answers to our planning questions so they can guide us in our execution.

This will be a month of intense work on moving content from the old site structure to the new. This will include the population of "contributions" for activities and events, including the creation of contributions for important past events and dormant activities that should nevertheless have a presence on the new site. Everyone on the team will have tasks to accomplish to meet the agreed upon timeline. At this stage the content will be created in a staging area rather than in public view. All team members, managers, and I will need access to this staging area.

Certain other technology options would also be nailed down during this phase. For example, an OR instance of WorldCat Local or a scripted "carousel" option would be chosen at this time.

During this phase check-in would be by phone, though at the end of the month the team would benefit from a face to face meeting to evaluate status and prepare for the next phase. By the end of the month we should have a relatively complete site staged, though I anticipate that some content may be quite compromised yet. For example, rather than a full-fledged WorldCat Local instance we may have to accept the flat publication listings for the initial rollout of the new site.

Deliverable: Staged site.

Deliverable: Working carousel.

Deliverable: Working micro-update feed.

Timeframe: July

The Cleanup Phase

While entry of some of the contributions would continue into the cleanup phase, the site would be well enough filled in to benefit from some internal critique. I would meet with the team in Dublin to evaluate progress in the execution of the site and prepare for an internal unveiling. While we gather feedback from the OR staff we would also begin the task of moving the staged material into public view. Given the limited time available, we cannot promise too many adjustments in response to staff feedback. Simple fixes will be made, but more involved issues will be added to a list of changes to be made as time allows after the public rollout. We will also create a rollout checklist that describes the steps we have to take to turn on the new site and turn off the old sites. This process may be done over the last couple weeks of August, though most of this work will happen, of necessity, on the day of the launch.

Deliverable: Rollout checklist.

Deliverable: Internal survey.

Deliverable: List of recommended future enhancements.

Deliverable: Public web site.

Timeframe: August

Some questions for the Planning Phase

These are an assortment of the kind of questions we need to answer in the planning phase. This is by no means a complete set, it is just meant to demonstrate the challenges of the task.

What can serve as our staging area on Collage? Should we use /research even though that is part of the current public site? Would it be easier to use /research2 and then swap it out for research when we go public?

How can we make certain pages show up in the Collage breadcrumbs even though we want them invisible to the Collage sidebar?

What can we use as a technology for the carousel? Can we be sure this will work with Collage?

Who can help us assemble some nice images to use on the site?

What elements should be included in the "contribution" templates for activities, people, and events?

Can we bet a WorldCat Local instance for OR? What portion of our publications would then need retrospective cataloging in order to be accessible via WCL? Can we drive the display of subsets of publications from WCL?

Will the simpler list for ResearchWorks be acceptable to Lorcan and the research scientists? In particular, will they accept moving "learn more" links to the activity page rather than having them available on the list of activities?

How can we structure and harvest Twitter logs so that they fold into our micro-updates? How can we flow news and blog posts into that same stream? What options do we have for displaying the resulting feed in Collage?

How critical is persistence of old URLs? Can we identify sections of the old site for which such persistence is more critical than others?

Concerns

I would need full access to Collage from Minnesota. This may require access to OCLC's VPN or some other kind of account. Proper supervision and participation in the work, though, would definitely require such access.

Time will be very much at a premium. Some questions for managers like Lorcan and Jim will need quick responses so that we can keep momentum on the implementation. In some cases we may be requesting compromise on sensitive issues, such as the content of the ResearchWorks page. We are planning this project for a summer and the managers are very busy people, so I imagine this attention will be hard to come by.

Summertime may also make it hard to get the attention of the team. I have no idea what sort of vacation plans team members have, and no downtime has been accounted for in the timeline. The team members also have full-time jobs already, yet this task cannot be completed on such a short timeline without some extraordinary commitment by the team to the tasks at hand. We need to make way in everyone's schedule for this implementation.

The communications staff (Bob and Melissa) and Patrick, in particular, will need to be forgiven other obligations so that they can move this agenda forward. These team members will have to also participate in online conversations and the maintenance of Q&A pages, templates, and other documentation of the project. I did not see this level of commitment during the wireframe development and am concerned that a lesser effort cannot meet expectations.

Operating on such a tight timeframe may require some severe compromises. Are we really ready to let go of some convictions? Persistent URLs? Database-driven publications? We don't have much time to develop complex solutions, and pressing forward may require that far from being ingenious, we may need to take shortcuts. Are we willing to accept that?

Timeline

The three months of this project would be June, July, and August.

• 5 June: Kickoff meeting with OCLC staff: developing questions to be answered during planning. Assign content inventory work to team (Bob?). (via phone)

• 17 June: content inventory with destinations completed due.

• 17 June: Meeting 2, assess template drafts, review Q&A progress, draft implementation timeline. (face2face)

• 29 June: Meeting 3, finalize plans and transition to execution. Possibly some participation of Jim and Lorcan so they can be fully appraised of compromises we anticipate. (face2face)

• 30 June - 27 July: Intense work building the site. Team moving existing content, populating Collage "contributions", adding data to WorldCat. I will be working on carousel and micro-update feed. Communication via email and as-needed phone calls.

• 20 July: Meeting 4, evaluate execution and transition to cleanup tasks. Review content inventory to ensure everything is moved. Develop final rollout checklist. Develop internal survey instrument. (face2face)

• 31 July: launch internal survey of staged site.

• 10 August: Meeting 5, any remaining execution tasks should be complete, evaluate internal feedback, begin working through the rollout checklist. (face2face)

• 28 August: Meeting 6, resolve last issues, create list of desired changes to guide future work. Complete rollout checklist. (face2face)

• 31 August: launch new site, remove old sites.

Costs

This implementation process will require a significant degree of presence either in Dublin or for long meetings on the phone. I've tried to minimize the costs, but there is no getting around the fact that we are trying to pack an intense bit of work into these three months. The total cost of this project will be $15,000 plus whatever expenses there are getting back and forth to Dublin. Given the range of airfare I saw during the spring effort, I'd estimate an additional travel cost of $500 per face to face meeting. I have arranged the meetings in the timeline so that I could be present in Dublin for all of them, though I think if we need to economize we could probably manage with half these meetings handled remotely. That would add an estimated travel cost of $1,500 to $3,000.

If this cost is too high, we could remove some of the technology identification and development from my plate. If OCLC could identify other resources for digging up a reasonable carousel option or micro-update facility, we could remove $3,000 from the total.

Assuming the project as described above, I would propose to invoice $5,000 at the end of each of June, July, and August.

Eric Celeste

Eric brings over 15 years of library and 25 years of technology experience to his consulting. At MIT Eric shepherded the creation of DSpace, open source digital repository management software developed with HP and now deployed at hundreds of institutions worldwide. At the University of Minnesota Libraries he encouraged the development of the UThink blog service, a wiki-based staff intranet, LibData, and the University Digital Conservancy. He works with non-profit institutions on appropriate uses of technology for informing, communicating, and collaborating with their constituencies.

/wiki/oclc/implementationproposal