Business Intelligence
at Indiana University

The Office of Research Administration has developed a ‘Digital Dashboard’ to provide fiscal officers and departmental account managers with revenue, expense, budget, expense forecast, and residual forecast  information which will facilitate the operational decisions and manage accounts. While this tool was developed specifically for ORA, expectations are that the functionality can be extended to benefit other departments, units, and campuses. The design was based on what the ORA fiscal officer needs in order to facilitate the management and reporting of the fiscal state of the organization. The Digital Dashboard reporting tool is being offered as a model to be used for an IU Business Intelligence Initiative Project that will have institution-wide value.

Key Aspects: Gain experience with SQL Server and OLAP (OnLine Analytical Processing), benefits a wide range of users, provides capabilities not currently available in institutional reporting systems

We continually face new challenges in enrollment management as we strive to:

• Offer the right courses at the right time to maximize enrollment, and
• Insure that students are able to enroll in the classes they need to satisfy their degree requirements in a reasonable timeframe

Enhancing our ability to accomplish this will have a positive impact on retention, graduation rates, student satisfaction, tuition revenue, recruitment and more.

We currently do not have the enterprise-wide information and reporting tools needed to sustain improvement in this vital arena.

The primary objective of this project is to provide academic units with integrated information and a variety of powerful tools which will empower enrollment managers and enable them to make the sound, expedient decisions needed to optimize enrollment management.

Key Aspects: Integrate data across applications and modules; provide information geared towards management; utilize metrics; continuity of data; tools and logic institution wide

This project begins by exploring the feasibility of providing a single BI solution for the reporting of the enrolment activities across the university including applicants, admits, deposits, and enrolled students.  Currently, at least four entities generate similar reports but they do vary in terms of timing, logic, audience and content. Consequently we find ourselves in an environment where multiple resources are performing a similar function, information is being downloaded in multiple areas, logic may not be consistent across areas and significant time is spent reconciling reports.

The primary deliverables of this project would include a centrally located single source for information based on consensus logic as well as a set of tools and reports that can be used by all campuses and departments to analyze, present and distribute the admissions information needed to support operations, external reporting, decision making and executive reporting.

Key Aspects: Consensus logic; starting point for student life cycle; single data source and tool set

The importance of Student Retention has been embraced system-wide at IU, and clearly impacts student satisfaction, alumni giving, tuition revenue, course management, national rankings, etc. The offices of the Registrar-Bloomington and University Reporting and Research have partnered to create a flexible system which will serve as a data source for tracking populations of undergraduate students. This system offers consistent definitions, consistent methodologies, less replication of work, the ability to develop comparison groups and facilitates coordinated retention efforts.

The focus of this project is to provide a set of BI tools to set on top of this data set that would make the information accessible to a broader range of users and also improve the consistency of analysis.  The end user will greatly benefit by having a sophisticated analysis tool, an interactive dashboard, drill downs into various populations, and devising comparison groups and consistent research approaches.  

Key Aspects: Leverage existing data; demonstrate use of a variety of BI tools; test bed for BI Tool Review project; provide value for a wide range of users

The objective of this project is to deliver a set of summary time trend statistics, integrated across a wide range of information domains including undergraduate and graduate admissions, student financial aid, student enrollment, course enrollment, instructional effort, staffing, budget, sponsored research, space, endowment, alumni and international activity.

BI Tools will provide users with time trends related to a set of summary activity indicators within each domain with the ability to drilldown to the program level and aggregate up through divisions/schools, campuses and the university as a whole. In order to facilitate analysis across information domains the system will include a “default” organizational mapping scheme.

Consequently this project has three primary components which are selecting the indicators, developing the organizational map and developing the actual application.

Key Aspects: Consolidation across modules and applications; development of institutional metrics; management level information; continuity of organization hierarchies; user-familiar tools

The objective of this project is to evaluate a variety of BI reporting tools based on a wide range of selection criteria and recommend a set of BI tools which will meet our needs and be supported on an institution-wide basis. A group made up primarily of analysts and end users has been formed to assemble the list of tools to review as well as the evaluation criteria. In addition UITS is providing an environment for the group to use during testing and evaluation.

Each tool will be classified according to their fundamental capabilities including Ad hoc Reporting, Analytics, Dashboards, OLAP and Report Development. Each tool will then be evaluated based on the criteria identified by the Tool Evaluation Group. It is not expected that we will find one tool which adequately fulfills all of the requirements for every project. Rather we will provide a “toolbox” of centrally supported BI Tools that will be utilized as needed to meet requirements specific to each project.

Key Aspects: Significant user input in selection process; not restricted to a single tool or vendor; testing and evaluation performed against functional BI project data

Our existing data warehouse was designed primarily for operational reporting. We have not yet implemented data models and tools designed specifically to support analytic BI which typically depend on very fast access to integrated summary information from a variety of applications and data sources. In order to realize the benefits of BI we will need to model our data differently.

Over the past two years a group of IU staff members were trained on Business Intelligence and data modeling. Since then a working group was formed to apply the information and techniques learned in the class, develop skills, identify standards and procedures and create shared information structures (dimensions and facts).

The work of the Data Modeling Group forms the foundation upon which we will build a data warehouse designed specifically for BI utilization.

Key Aspects: Use the foundation that exists; every project can add to the foundation; single source of data; centralized, shared business logic; enables BI

As an enterprise data warehouse, it is only natural that the amount of data contained in the data warehouse would continue to grow. The nightly refresh of the data (the ‘race to daylight’) consists mostly of full rebuilds of data warehouse objects which are based on application data. As the amount of data increases, the time required to update the warehouse also grows while the available time to complete the refresh remains static. Load processes must be streamlined to ensure data are refreshed and made available to the University community in a consistent and timely manner.

While no BI working group has been created with this as a focus there are several groups addressing this issue on a continuous, ongoing basis. One of the challenges for the BI Initiative is to increase the availability of information without extending the nightly build times. The BI Project Groups are working closely with the EDSS team in considering the potential impact and possible solutions to avoid a negative impact on an already limited refresh window.

Key Aspects: Maintain level of service to operational data warehouse; plan for additional needs; will benefit from data modeling long term; consider alternatives to full nightly builds; project teams partner with EDSS

Implementing BI Tools involves much more than providing developers with the means to create new reports. There is much to be considered from an infrastructure standpoint which must be considered when evaluating the cost of hosting and deploying new tools. Some of the associated costs may be related to new hardware to store the data or host the tool, or the human resources need to maintain systems or integrate tools into our environment (IUIE, OneStart, etc.).

While no BI working group has been created with the overall design of this architecture as a focus there are several systems and groups performing similar functions on an ongoing basis which are involved. The BI Project Groups are working closely with the EDSS team in considering potential strategies for implementing, integrating and supporting the initiatives.

Key Aspects: Objective is to get the most out of existing resources; project teams partner with EDSS;

There are a variety of aspects within all projects which should be approached consistently with some level of defined processes and standards including requirements gathering, metadata, data quality, testing, security, training, education, communication and resource estimation.

The development of our BI processes and standards will begin with a relatively simple outline of what should be considered when addressing different aspects of a project. As each project proceeds, the processes and standards will be further refined, making them more useful for subsequent projects. One example is requirements gathering. Each project will be different in this respect but there should be a standard set of issues addressed and questions to start with in order to avoid omissions and extra work.

Key Aspects: Projects start with a checklist primarily to insure all aspects are considered; each project contributes; refine over time; reduce redundant activities