On April 1 we reached out to the MOR Leaders alumni on behalf of Ed Clark, fellow program alum and current CIO of University of St Thomas, with a survey on "IT Centralization and the Innovation Value Chain in Higher Education". This was part of his PhD dissertation work, in which I am happy to report he passed and earned his degree. Congratulations Dr. Ed! As an expression of appreciation, Ed has drafted a summary of his findings to share with you all. Below please find that output.
A long time ago, when I was a fairly new college IT leader at the University of Minnesota, I ran into one of the most interesting—and frustrating—problems of my professional career. I was fortunate to have a talented IT team with many ambitious and brilliant individuals, and we had a reputation of being able to build and adapt tools to meet the emerging instructional and research needs of our departments and faculty.
One of these inventions, a tool called Media Mill (created by Colin McFadden, a truly great technology innovator), proved so popular with our faculty that it was quickly shared via word of mouth with other colleges. We learned that students and faculty across the university, and even outside the university, were using the tool to create and store media files and share them with other collaborators.
My budget at the college level couldn’t afford the storage and support costs of managing a large-scale service with rapidly expanding adoption. With this knowledge, I went to Steve Cawley, the CIO of the university, and asked him to consider taking over the application as an enterprise offering. While Steve was initially receptive to the idea, his team presented a host of doubts and concerns. The tool was written using Cocoa and Objective C, two newer (and relatively untested at the time) platforms. It ran on Xserve Apple servers, which were incompatible with the skills and environments deployed by the central IT office (OIT). I was told that because of these concerns, OIT could not take on the Media Mill service despite its widespread campus adoption. But the story wasn’t over.
1. The Innovation Value Chain
Of course, what I had experienced was an example of the many barriers that an organization faces during the innovation process. Ideally, new ideas are generated to solve strategic needs, the ideas are then converted to services and offerings, and then these services and offerings are diffused throughout the organization. This process is known in academic literature as the “innovation value chain.”
Many things can emerge as barriers and enablers at each phase of this process. Is the idea compatible with the organization’s existing norms? Is the invention easier to use than what it replaces? Is it harder to manage and support? In not considering the overall continuum of stakeholders and their needs, it was easy for me and my team to complain about the stubborn intransigence of the OIT team. However, was it reasonable to ask an enterprise team to change their current approaches and support the Cocoa platform and Xserve servers? Probably not.
On the other side, it was easy for some of the OIT staff to complain about the mavericks on my team building crazy, unsupportable things. However, were these teams thinking about the potential benefits for better teaching and learning outcomes? Did they consider the implications of the fact that new users loved the system enough to adopt it quickly and widely? Probably not.
2. From Moore’s Law to “More And More”
I am sure that every one of you has seen some version of this story play out at your respective campuses. But we are at a time where striking a balance between innovation and efficiency has never been more critical. My friend John O’Brien, President and CEO of Educause, recently gave a talk at Macalester College in St. Paul, MN. During his presentation, he made the following statement that really stuck with me:
“We’ve gone from Moore’s Law to ‘more and more’.”
If you think about it, Moore’s Law is a concept that means very little to most technology consumers. (Imagine explaining it to your children, or your parents and grandparents.) It is a metaphor for IT as a disconnected organism that keeps advancing for no specific end. This is the old “genius in the tower” (or “geek in the cave”) model of IT.
“More and more” on the other hand, is about that fact that the phone in your pocket right now connects you the collective knowledge of the world. If you are a student, you get your course materials and grades online. If you are employed, your job searches probably go through LinkedIn or Indeed.com. And Google and YouTube can help you with everything from cooking bok choy to determining whether your sore throat requires medical attention. This IT model is about shared expertise and exploration, or “shared leadership.” Everyone can contribute, and already does so at an unprecedented level.
For higher education, “more and more” also means that the federal government and the legislatures of most states want big changes in colleges and universities. They want better student outcomes, better retention, better 4-year graduation rates, and lower debt loads at graduation. Oh, and they want all those students to be employed at graduation, too. In short, they want higher education to innovate, and to do it quickly.
3. …With Less and Less?
Those of you at state institutions don’t need me to tell you that your state is investing in a smaller percentage of your school’s budget than it was 30 years ago. Furthermore, most of us—public or private—have been asked to become more efficient with our existing resources. Consulting firms like McKinsey or Huron are brought in at rapidly growing rates to identify efficiencies. Their focus is usually on operations: IT, facilities, human resources, auxiliary services, etc. The resulting IT recommendations generally revolve around centralizing staff, cutting redundant staff and services, and adopting common technology platforms and approaches.
4. The Study
As we have moved from Moore’s Law to “more and more,” under a financial environment of “less and less,” it was important to me to research how shared leadership (or its opposite, centralized leadership) affects the innovation value chain in higher education. As I worked on my dissertation at Mankato, I had the opportunity to survey many of you MOR participants to measure the levels of shared leadership at your institutions, while also measuring the strengths and weaknesses of your respective innovation value chain processes. Briefly, shared leadership was a measure of: 1) to what extent were all IT workers (centralized and decentralized) involved in central IT decisions at an institution, and 2) to what extent were central IT approval processes a hurdle to trying new things. Correspondingly, the innovation value chain questions and answers reflected to what extent there were problems in idea generation, idea conversion, and idea diffusion at each institution.
The results are in! Here are some of the most interesting results (look for more detail in an upcoming Educause Research article, and even more details in my dissertation on Proquest):
a. Private institutions were hurt by lower shared leadership in all phases of the innovation process, from idea generation to idea conversion to idea diffusion. These impacts were all “medium” on the Cohen effect size scale. To put that in perspective, Cohen describes a medium effect size as “large enough to be visible to the naked eye,” citing the difference in height between 18-year old women and 14-year old girls. Another medium effect size example is the correlation between education levels and incomes for English-speaking workers in the United States. For topics like innovation, findings at these levels are a big deal.
b. Furthermore, decentralized IT staff at private institutions reported results that showed large adverse impacts of lower shared leadership scores on the idea generation phase. Cohen describes large impacts as “grossly perceptible and therefore large.” Examples of large correlations include the relationship between increased cigarette smoking and increased likelihood of lung cancer, or the connection between higher verbal SAT scores with higher math SAT scores.
c. Joint reporting IT staff, those that reported to both a local unit and the central IT unit at their institutions, reflected large adverse impacts of lower shared leadership on idea conversion (the stage where ideas are converted to new services and offerings). This was true for joint reports at both private and public institutions.
d. Finally, public institutions were not immune. Low shared leadership had a medium adverse effect on idea generation at public universities, and a low-to-medium adverse effect on idea conversion. Interestingly, idea diffusion was barely affected at these schools. This difference in results between public and private universities should be studied more thoroughly.
So how did the Media Mill story end? Ultimately, Steve and OIT agreed to collaborate with my team to run Media Mill for the University of Minnesota. They also agreed to fund the hardware and storage upgrades that would be required to meet the needs of an enterprise level service. What a great solution to a thorny problem! We would share the leadership of delivering an important innovation to the campus. Over time, Media Mill was upgraded to a more sustainable architecture and the keys were handed off to OIT (it is still being widely used at the University of Minnesota today).
While there may be limits to what you can do to reverse cost cutting measures at your institution, there is still a lot that you can do to promote shared leadership in your IT organizations. Shared leadership is something you have learned from the beginning in the MOR program, from coaching sessions to group interactions to building your relationships with others. More importantly, shared leadership is something that your institutions need to champion in order to assimilate innovations more readily. It is the key to doing “more and more” in an environment of less and less.