All Aboard the Training Train – Keeping your Business Intelligence (BI) Solution on Track:
Jamie Sweeney, University of Texas
You have the data warehouse and BI solution – now what do you do? This is not a “build it and they will come” approach. Technology alone will not sell itself. No matter how wonderful the BI solution, users will not use it if they do not have trust in the data or feel comfortable using the tool. Invest in a quality training and outreach program – blow their socks off with training! Get users excited about going back and using the tool immediately! After training build relationships by reaching out to the campus community to gain customer feedback to help enhance and improve upon your existing tools, support, and service. This presentation will give attendees an overview of the training process start to finish, noting possible bumps along the way, and how a good training and outreach program can keep your BI solution on track. All Aboard . . . (Presentation Takeaways) Helpful Tips for: – Developing your Training Program – Creating an Effective Training Experience – Building Customer Relationships through Outreach – Measuring Success.
An EPM Deployment Case Study:
Shahriar Panahi, University of Massachusetts
The University Of Massachusetts is currently implementing Oracle’s Enterprise Performance Management (EPM). We see a lot of potential for using EPM as the foundation for the University’s Enterprise Data Warehouse, and are moving forward with implementing it. Further, we are using OBIEE as our primary access layer solution and are implementing some of the Oracle delivered Fusion Intelligence. This presentation will review our objectives with EPM, enumerate some of the implementation challenges and offer a list of lessons learned.
An Intuitive Data Mart Built with Confusing Tools and Technologies:
Daniel Riehs, Boston College
During the past year, Boston College has been in the midst of an extensive effort to improve its data warehouse by building two new data marts, each with specific business users and purposes. A goal of the project was to construct intuitive systems that would not require users to be familiar with technical data warehouse concepts, such as fact and dimension tables. This presentation will detail some of the strategies used to make this possible, as well as the obstacles that were faced. Special attention will be given to situations where we were impeded by the tools used to construct the data marts, as well as the tension discovered when functionality that should be driven by business users, such as meta-data and documentation, is included in applications created for technical audiences. The presentation should appeal to anyone interested in developing well-documented reporting environments, but particularly helpful for those using Cognos and FrameWork Manager.
BI Solutions for Sponsored Research Reporting and Analysis:
Aaron Waltz, University of Illinois
Co-Presented by:
Beth Ladd, University of Illinois
Effective information solutions to support sponsored research are more important than ever in today’s climate of declining budgets and changes in regulations governing grants administration, such as the American Recovery and Reinvestment Act (ARRA). However, grants reporting poses a number of challenges. Grants data may reside in multiple systems. Information needs are diverse and span different types of customers. Grants data is fairly complex to use and understand, especially for business users and faculty researchers who often do not have financial expertise. To address these challenges, the Decision Support department at the University of Illinois created a set of solutions for Grants reporting and analysis. A product plan was created to define the primary types of information needs, data sources, data marts, and what BI capabilities would offer effective solutions. The product plan was also used to identify and scope a set of projects that could be executed separately but that resulted in building a comprehensive, interconnected solution. This includes a set of grants data marts, with ad-hoc query and analysis. Additionally, prototyping is underway for grants dashboards to provide graphical summaries and trend information, as well as an OLAP cube to support dynamic analysis of the grant lifecycle.
Building and Scaling Data Warehouse and Business Intelligence as Enablers of University Business:
Daniel O’Connell, Yale
Co-Presenter by:
Hans Son, Yale
How to plan, design and build an enduring and scalable Data Warehouse and Business Intelligence Architecture enabling the University Business by leveraging Database/Data Warehouse machines/appliances, in-memory databases and enterprise Business Intelligence Solutions such as Oracle BIEE Plus, MSSSRS, MSSSAS, Informatica, ODI, Essbase, Exadata, TimeTen Database, Teradata, Netezza and other technologies.
Building a Productive Relationship between ETL Analysts and Source File Stewards:
Rainbow Di Benedetto, University of Texas Austin
In this presentation we will address the issue of cultivating a productive relationship between ETL analysts and source file stewards. We will discuss the topic from the following three perspectives: (1) why a productive and constructive relationship between these two parties is critical; (2) how we have been handling it here at Information Quest at University of Texas at Austin with both our best practice and lessons shared; (3) how our experience has fostered a strong cooperation and mutual understanding and presented win-win solutions for both parties.
A Comprehensive Dimensional Data Model for Enrollment Management:
Bob Duniway, Seattle University
Strategic enrollment management (SEM) is critical to the success of colleges and universities. Given the complex set of interdependent issues involved in SEM, having an integrated data model around which to build BI support makes sense. Unfortunately the complexity of SEM makes developing such a model unusually challenging, and most institutions build BI tools to address only a subset of the issues. This session will present and explain an integrated dimensional model institutions can use to organize BI support for SEM.
Controlling the Chaos: Storing Business rules in a Database:
Christine Ray, Indiana University
Business rules for organizing data in a Balance Sheet, Income Statement or a Cash Flow Statement may be put into a central location, database, or into each and every report. It makes sense to put the rules into the database instead of repeating the rules, sometimes in many different reporting languages, throughout the multiple reports. With an ETL tool you can take the simple rules and expand them to add the necessary classifications and sort orders to the General Ledger Balance and/or Detail tables. Putting the rules in the database will allow a report writer to write many different styles of each statement without having to duplicate the rules. I will show how rules can be implemented in the database, expanded to the GL and/or detail tables and used in a data dump or reporting environment. We have a variety of report needs within each financial statement and having the rules in the database allows for a single location updating. A short demo of the resulting financial statements will be shown using BIRT (Business Intelligence and Reporting Tool). BIRT is an open source software project providing reporting and business intelligence capabilities.
Creating Decision Centers:
Marco Cestaro, State University of New York
Bridging the gap between BI Professionals and non-BI professionals is one of the greatest challenges of any Business Intelligence program. Decision Centers create a conduit that can be used by those in need of information to communicate to those who provide the information. This creates a link between business strategies and tactics, the information required to support them, and the technologies used to deliver information. This session will be an introduction to Decision Centers and the components that comprise them. An expanded review of the methodology used to develop Decision Centers is the focus for much of the session. A brief look at where this process fits into an overall strategic plan for information access will be included as well.
Cubes are Not Toys: Fast-track to Dimensional Analysis:
Nancy McQuillen, University of Washington
Cubes provide a viable means of delivering data to end-users as an alternative to writing complex SQL queries. In 2009 the University of Washington delivered a production cube of the prior biennium’s financial transactions, supporting ad hoc questions about revenues, expenses, and other measures. The financial measures can be summarized by any combination of 12 built-in dimensions to answer questions about total dollars by organization, by account, by budget, by program, by job classification, etc. Using Excel 2007 pivot tables as the client tool, the cube was used to assist in summarizing measures for the university’s annual report, and to answer a variety of financial auditors’ questions. The presentation will provide an overview of this Microsoft SQL Server Analysis Services cube design, plus explanation of how the technology enabled delivery of current-state data warehouse data into dimensional form: • without ETL development, • without completion of master data projects and finalization of conformed dimensions, • without implementing a dimensional model in the relational database used as the cube source, • without requiring SQL queries by data end-users, and • without extensive training for users already familiar with Excel. The session will emphasize generic features of cube technology vs. vendor specifics.
Data Warehouse-Based Decision Support for Higher Education Management:
Ellen Hoeckmann, University of Osnabruck (Germany)
Co-Presented by:
Sonja Schulze, University of Osnabruck (Germany)
This presentation illustrates the implemented concept at the University of Osnabrueck using a flexible, analytical reporting based on a data warehouse to support users’ strategical and operational decisions effectively. Increasing (inter)national competition in higher education requires efforts of all university members to improve personal and organizational performance. We will present decision support applications for all kinds of members, ranging from executives to faculty, including students as well, and even first and foremost. Originally started as a pure reporting project in 1998, R&D activities were soon redirected towards decision support for university core processes due to user requirements and acceptance issues. Functions include benchmarks for students to monitor personal performance, exams scheduling using heuristic optimization for faculty members, balanced scorecard based performance monitoring and redesign of study and research programs for campus executives. Actually, users include about 10000 students and 1200 faculty members in 175 programs of study. The overwhelming success mainly results from the bottom-up strategy concerning the supported user levels. Critical success factors are early and continuous involvement of users, supported by an interactive web-based user interface.
Developing Key Performance Indicators:
Insiyah Jamal, Drexel University
The second phase of Drexel University’s BI initiative for Enrollment Management focuses on implementing key performance indicators. Attend this session to learn about: – the process of identifying measures to be tracked – KPIs implemented for the Admissions and Student areas – a short demo showing the KPI interface functionality and the underlying reports.
Getting the Keys to the Kingdom:
Eda Matthews, University of Texas Austin
This presentation is geared towards those members interested in application security for their BI initiatives.
Hybrid Data Modeling:
Madan Dorairaj, Princeton University
This session is intended to provide an insight into Princeton’s approach in building a data mart that would serve two purposes. Firstly, there is an immediate need for “list” reporting that will help create satisfied user community. Secondly, there is also an appetite for data driven decision making type reporting. Needless to say, “list-type” reporting provides quick ROI and helps secure confidence and necessary support for implementing “explorer-type” reporting. The challenge is to weigh in both needs, the former is explicit and easily understood while the latter is not so, and design one model – hybrid model, that will meet the needs of both user community.
Implementing OBIEE at Cornell University:
Stephanie Herrick, Cornell
Co-Presenter(s):
- Neil Belcher, Cornell
- Elinor Poole, Cornell
- Judy Kany, Cornell
Cornell University is in the midst of implementing OBIEE as its enterprise BI product. This presentation will give an overview of Cornell’s OBIEE Project as well as a demo/walkthrough of OBIEE reports & dashboards, ad hoc Answers, and the OBIEE repository as they are being implemented at Cornell.
Institutional Research – Evolution and Transformation of Campus Information into a Fully Integrated Decision Support System:
Kimberly Register, University of California Santa Cruz
Co-Presented by:
Kathleen Dettman, UC Office of the President
The recently established Institutional Research Office at the University of California Office of the President (UCOP) is charged with transforming multiple data repositories (representing data collected from multiple campuses & research locations, each running disparate ERPs) into an integrated and broadly accessible Decision Support System. This large-scale effort will build on the demonstrated success of the University of California Santa Cruz’s (UCSC) data warehouse. Over the last 15 years UCSC developed a mature data warehouse that integrates information from its key business systems, while making information readily accessible to its constituency. This presentation will explore how the UCOP effort will benefit from UCSC’s experience with: 1) using a component-based approach to effectively integrate campus data and develop a robust reporting environment, and 2) leveraging constituent participation at every stage. A component-based approach to development identifies key elements of a reporting data set that can be redeployed in subsequent efforts, reducing implementation time and assuring linkages across data sets. Actively engaging campus constituencies in the full cycle of development ensures “getting it right the first time”, and best utilizes limited technical resources while building a broad base of support.
Keynote – Data-driven Decision Making
Jack Maguire
Organizational Resistance to Data Warehouse Programs:
Bruce Jenks, Golden Gate University
Where does resistance to data warehouse programs come from? How is it manifested? What have people done in response to it? Based on more than twenty in-depth interviews with data warehouse consultants, IT professionals, and power users from higher education and commerce, I will present my preliminary findings against the backdrop of organizational change. Initial data indicate some consistent patterns for both the antecedents and consequents of organizational resistance to data warehouse programs. Of particular interest are the similarities and differences between higher education and commerce as seen from the role-based perspectives of external consultant, IT professional, and power user. The data to be presented was collected as part of a dissertation research project.
Panel – “I Don’t Get No Respect” Turning Data Warehouses Into Campus Priorities:
Ted Bross, Princeton University
Panelists:
- Liz Wallace, American Public University
- Ora Fish, NYU
- Daniel O’Connell, Yale University
- Ian Wall, Harvard University
This panel discussion will look at how 4 institutions brought their Data Warehouse initiatives to become a high priority on their campuses. Too often, the reporting and warehousing functions are afterthoughts to ERP/Transactional systems implementations.
Panel: Leading an Effective IR-IT Partnership
Bob Duniway, Seattle University
Panelists:
- Anja Canfield-Budde, U. of Washington
- Rainbow Di Benadetto, U. of Texas at Austin
- Kathleen Dettman, U. of California Office of the President
- Michael Glasser, UMBC
- Emily Thomas, Stony Brook University
Panel: Lessons Learned in BI Governance:
Catherine Lloyd, Harvard University
Panelists:
- Aaron Walz, University of Illinois
- Anja Canfield-Budde, University of Washington
- Ted Bross, Princeton University
Join our panelists in a lively discussion of BI governance at their institutions. Session will cover:
- Mission, vision, strategy, and organization of BI program
- Executive leadership and stakeholder engagement
- BI initiative prioritization and approval in constrained economic times
Performance Management on a Shoestring: A Case History:
Andrea Ballinger, University of Illinois
Many organizations use performance management techniques to set strategic goals and then measure the progress of the organization toward those goals. In these times of tight university budgets getting a performance management initiative started is especially challenging. Many consultants and companies offer services and training but at a significant cost. In the role of a Business Intelligence Competency Center, Decision Support at the University of Illinois has advised units interested in performance management. However, when Decision Support staff began their own performance management effort, they immediately brainstormed over 100 key performance indicators. Before actually beginning design, the same staff realized that they needed to start at the beginning and update the Decision Support strategy map. Over the last nine months they have used internal resources only, free webinars, white papers and a very minimal budget. This presentation will address the challenges the Decision Support team encountered and solutions they have adopted. While the project has several more steps to complete to field its first measures, the team has insight to share in terms of the resources they’ve found and lessons they’ve learned.
The Rapid Response Data Mart:
Rachel Gatlin, University of Washington
Co-Presented by:
Joseph DeVore, University of Washington
The American Recovery and Reinvestment Act (ARRA) introduced stringent new reporting standards for research and projects funded by stimulus funds. The pressures of ARRA’s timelines and requirements threw out normal planning practices and required a rapid response. In under four months the UW developed an incredibly successful data mart based system for reporting that reduced the required human effort by over 98% compared to a proposed manual solution by combining data from our central DW with transactional data from areas not yet incorporated into the central DW and then building application and reporting interfaces. Presentation will include design, challenges and decisions made in the creation of the ARRA Mart, the outcomes (foreseen and unforeseen), the nature of the collaborative effort across business and technical units that had not previously worked together, and the new opportunities to improve our transaction systems, our business processes and the build out of our central DW including improvements to an area currently in the process of building out in our central DW. Our “Rapid Response Data Mart” for ARRA demonstrates the value of using all available information to enable agility and efficiency in a time of increasing regulation and decreasing funding.
Relational vs. OLAP in the Dimensional World:
Steve Grantham, Boise State University
Dimensional data models can be implemented within either a relational database or an OLAP cube architecture; each has its own strengths and weaknesses. To a first approximation, the relational layer is sort of like a well used 4WD pickup truck; you can get almost anywhere with it, and carry almost anything, but it may not be the fastest way and the ride may be a bit bumpy. The OLAP layer is more like a sports car: on the right roads it’s faster and more fun, but it takes a bit of maintenance to make it run right, and there are some places you just can’t take it and some things you can’t do with it. In this presentation I will illustrate this analogy with some specific examples of various types of reporting needs and which layer supports each one best. This will be based on Boise State’s implementation of the iStrategy student data warehouse, but I believe the principles illustrated are fairly general.
Rensselaer’s Data Warehouse: Rollout, Training, and Culture Shock:
Keith Cushing, Rensselaer Polytechnic Institute
Rensselaer Polytechnic Institute improved decision support and changed its administrative culture with an award-winning data warehouse. In this presentation, you will learn how RPI’s data warehouse accommodates the analytical, operational, managerial, and personal data requirements of the Institute’s diverse user-base. Attendees will see how users interact with the data warehouse through dashboards, dynamic reports, and ad hoc query tools. In addition, the presenter will describe how the data warehouse team at RPI works with campus constituencies to roll out data marts and dashboards, train and support end-users, and change the Institute’s culture. Whether you’re just in the planning stages or currently rolling out your data warehouse, you’ll gain insight from this presentation on RPI’s front-end data warehouse experience.
Technical Lightning Talks
Helen Ernst, State University of New York
Time Keeps on Slipping… Automating Report Date Filters through Metadata:
Darin Mattke, University of Texas Austin
Situation: Data is loaded at variable release dates (12th Class day; after ‘close’ of month). It is available in our reporting environment to a select set of users and data stewards for data validation prior to production release. Issue: Data is in a single location, i.e. we do not utilize a test database environment for data validation prior to loading to production. The difference between data validation and production reports is determined by date filter criteria hard-coded in each report. Process: Reports and cubes are automatically refreshed using a system of scripts and utilities. However, definitions of data validation and production reports are manually and individually updated. This opens the possibility of human error in a number of instances and is time consuming to re-code reports in addition to the routine data validation and verification process. Solution: “Chronos!” A home-grown initiative to automate the process of placing data into various states of production. A dynamic date dimension is used for a table-driven filter system that will identify valid date ranges and various readiness states of data. This reduces report coding time and the error-prone process. See how this is done and discuss alternative options!
Tuition Distribution at University of Pennsylvania:
Francesca Seidita, University of Pennsylvania
Co-Presenter(s):
Ed Stemmler, University of Pennsylvania
The design and implementation of a Dimensional Model used for communicating as well as distributing tuition revenue from a central Billing & Receivables (BRS) system to schools using a RCM financial system. Penn’s previous tuition distribution process consisted of manual calculations by the comptroller’s office using student registration counts and tuition income from the BRS system. The process was essentially black box, with little transparency. The new model, based on Kimball’s methodology, consists of a dimensional model which supports analyses by courses, students, instructors, across five snapshots each term. The registration data, tuition revenue, and distribution values are fully accessible by the financial administrators using Business Objects across the University.
What We Learned Building a Faculty Dashboard:
Matthew Sheppard, Arizona State University
After assuming responsibility for ASU’s Dashboard Division, I was immediately tasked with building a Faculty Dashboard. As the project manager for this particular dashboard I was involved from the requirements gathering and scoping phases to the delivery and roll out of this tool. I tackled complicated issues relating to the university’s data and how to bring various sources together to generate the requested reports and measures. I managed expectations and interacted with the lead customers from the Provost’s Office, and in the current political environment was forced to present and highlight only certain types of data related to faculty members. Additionally, I tackled issues relating to security in an interdisciplinary atmosphere where the directive was to only allow departments to view data they had a direct interest in. We’ve completed work on this dashboard and as we put the finishing touches on the security and prepare to roll this dashboard out across our campus I can reflect back on the valuable experiences and lessons learned to bring a tool like this into the culture of a higher education institution. I’d like to share my insight with others that stand to benefit from our trials and tribulations.
Wildcats and Sun Devils – Using BI to Boost BI:
John Rome, Arizona State University
Co-Presenter(s):
Doug Hester, University of Arizona
The University of Arizona and Arizona State University will put their in-state rivalry aside for the day and demonstrate how they capture usage and web analytic information from their BI environments (OBIEE, Hyperion, Corda, web, etc.). Armed with this information, these Wildcats and Sun Devils can now measure adoption, know their customers, and be proactive in providing a better experience and service. The presenters will show how they collect this information, share their stories, and list a few benefits they’ve seen.
Zen and the Art of Data Warehouse Management:
Sean Blood, University of Massachusetts
Many of us concentrate on the development and start of a data warehouse solution/program. This presentation deals with what happens after delivery – maintenance, governance, continued development, roadmaps, user expectations, etc. What worked and what didn’t and why. The goal is to discuss how UMASS did this while keeping our sanity. Hence the Zen part of the presentation.