Recently on Micronet, the mailing list used for discussion by the campus IT community, there have been various comments about the Adobe Creative Suite project and the process by which it was implemented that raise important issues regarding future OE-related IT efforts. In the discussion, there were several implied statements or questions pertaining to business authority (e.g., decision rights), control, and process failure/improvement that I'd like to address with the broader community.
Decision rights
In the case of the business authority to make technology-oriented decisions that have campuswide impact, governance is a key issue for the campus. Do we make decisions based on organization? On funding? On individual choice? Or perhaps decisions are based on which of those three will get us to the answer that best suits the individual situation. Under what situations do our roles change from having input rights to decision rights? We do not have a comprehensive IT governance model in place, the result of which is a lack of clarity about what decision path we should take under what circumstances. Purchasing software would seem to be straightforward but in fact, those decisions have historically followed all of the above-mentioned decision models.
To improve the situation the OE IT Governance project was established to create a framework on decision rights allocation and make sure it is consistently applied. That project is being lead by Liz Marsh in the Office of the CIO and will be working on an overall governance model for the campus, including a process to identify decision right areas that need clarification. The identified areas will be brought to the OE Executive Committee, which will vote to allocate decision rights as appropriate to the individual, departmental, campus, or systemwide level. In the case of the Productivity Suite, the decision rights to select and implement productivity software for the campus were assigned to the VC of Administration and Finance (John Wilton) and the AVC-IT & CIO (myself). If you have questions about this please feel free to contact me directly,
Control
Some messages implied that individuals in groups outside your own are making decisions they don't have the authority to make and implementing processes that are somehow designed to remove control from local IT staff. That type of thinking suggests a purposeful effort to make your job more difficult when nothing could be further from the actual goal of these projects. However, designing processes that work for a place as complex and fragmented as Berkeley is not an easy undertaking. Doing something one way for one group often means it won't work well for another without all of us agreeing to adapt in small ways to achieve the larger goals.
The Adobe contract, and subsequent ones that are soon to be released, each have implementation challenges related to software distribution, key management, renewals, support, patching, etc. Rather than delaying the release of the software, a decision was made to release the software as soon as possible to avoid expense to students, faculty, and staff, who would otherwise be purchasing at full educational list pricing. In the first two weeks of the deployment alone we avoided more than $5,000 in University purchases that were attempted after the program had started. It's not about control, it's about balancing the need to do things quickly and the desire to do them perfectly. Since the outset of this project, we have been, and continue to be, open to suggestions about how we can make this significant shift in how software is delivered at a campuswide level work best for everyone.
Process
There is no question that we all have room for improvement in the way we operate IT on the campus — at the department, control unit, and campus level. Our past approach of each person doing their best in their area results in some local optimization but little chance to achieve significant gains at scale. Full centralization may achieve scale but would result in significant degradation in our ability to meet the diverse needs of our very large community. So what is the answer? Work together and apply process changes to gain benefits and cost savings from scale in as many areas as we can while making sure to clearly identify and meet the local needs that are unique to a given discipline or department.
The challenge is that in our current environment, virtually everything "appears" to be unique because that is how we have always done it. Not because we should do it that way, or because there are no other options, but simply because that has been our past practice. Installing software on end-user devices is not a unique or highly specialized activity, but one we need to find ways to lower the cost of doing so IT staff can spend more time on the important value-added activities. To get there we need partnership, collaboration, suggestions, and patience as we work through building a new, more cost-effective model for the campus.
Our path forward is clear — we must find ways to keep the money going to classrooms and research, and reduce our duplicate spend in every area we can. The time is now to put aside our past differences and divisions and trust each other to be working toward that common goal.