Academic institutions all across the U.S. are becoming increasingly aware of the significant gap between their imbedded data and analytic systems and the growing Business Intelligence (BI) needs of their organizations. It is not the lack of information that has created this gap but the relative disjointed nature and taxonomy between the existing data and the decision-support systems available to university administrative and academic management.
Several historic circumstances have resulted in the current state. First, universities (particularly research-extensive institutions) tend to departmentally focus on software and data system implementations. Whether for finance, facilities, research, student services, academic operations or other functions, software/system tools were often selected and implemented for the specific needs of that operational area exclusive of institution-wide data expectations. Second, these implementations occurred over multiple technological time periods resulting in highly variable vendor solutions (i.e. software and operating systems). Finally, the data definitions and nomenclature used for these and other college or department-based systems are not consistent and therefore difficult to integrate into shared reporting services. This information ecosystem can result in highly variable reports, manual corrective measures, and general distrust of reported organization performance results.
However, higher educational institutions are discovering the crucial value of fully integrated and modeled data for business decision support. Today's competitive nature of academics (enrollment and faculty productivity), funding constraints, and student success metrics are some of the key drivers for this data governance awakening. The cyclical nature of the U.S. student population along with rapid online and international educational growth is resulting in the necessity for better understanding of course and program success as well as possible tuition price elasticity. Accreditation standards are also tied to these success factors. State and federal funding metrics are shifting from credit hours to student accountability and degree/career success. Additionally, the sources of data are expanding significantly with public/private cloud information (publications, presentations, etc) being added at a rapid pace.
So what are some basic steps academic institutions are taking to create a more comprehensive reporting environment? An excellent first step is to obtain buy-in to the general principles of data governance at the highest levels in the organization. This includes establishing transparency expectations, appointing key data trustees, creating a high-level process for development, review, and approval of data elements, and identifying the technology underpinnings for success. Organizations who initially establish this baseline of expectations (guiding principles) create the clearest path for success. This roadmap provides the data stewards and IT with both, the expectations and confidence to implement effective processes, procedures, and conflict resolution guidelines and policies.
“The university’s leadership should be made aware that robust BI and data warehousing tools are initially just empty containers requiring ground-up configuration”
From the IT perspective, inventory analysis of the portfolio of BI tools across an academic institution usually results in the discovery of several commercial and/or homegrown solutions. Staff allocation and niche knowledge/skills in both academic and administrative units are typically dedicated singly to these individual tools. It may be beneficial to refrain from initially imposing a mandate to migrate to a single centralized tool. If attempted, this can often result in end-user resistance and frustration before functional benefits can be fully identified and embraced by the institution. So IT may need to support integration of multiple tools and their data sources while ultimately pursing a consolidated and optimized set of tools and resources. Finally, this approach can lead to initial quick functional wins while setting a path for unified data service success. The “big bang” approach to data system implementation takes significant corporate will so consider a multi-year staged effort for lasting success.
One final step for consideration is determining the optimal IT and functional staff resource allocations. To a large degree, this area should be driven by substantive review and implementation of the data principles and processes described above. However, two additional factors are worth an institution’s attention. First, the siloed data approach described above typically indicates minimal functional and IT staff allocations currently exist across the organization. Otherwise adequate staffing levels would most likely have resulted in a more comprehensive data management environment. Second, the university’s leadership should be made aware that robust business intelligence and data warehousing tools are initially just empty containers requiring ground-up configuration. So an effective BI system typically requires more staff resources than anticipated and perhaps currently available within the institution.
It is clear this is a challenging area for higher education but also one with great potential for far-reaching benefits. Certainly many universities are already well on their way to successful data governance models and perhaps creating a competitive advantage over their peer institutions. Overall, academic institutions can no longer remain ambivalent to “Big Data” integration and analysis. The necessity of robust data analytic services to drive business decision-making is here to stay and effective future teaching, research, and service delivery will largely depend upon it.