A Concise Introduction to Analytics

A Concise Introduction to Analytics

Lixin Tao, Seidenberg School of CSIS

February 21, 2012

Quick Jump: Forum for Brainstorming, Forum for Identified Analytics Courses

A business or organization runs business processes to deliver products or services. A business process specifies the business’s input/output, workflow, decision making, and interaction with collaborating processes or environment. A meta-process is used to create new processes and adapt the existing processes for objective/environment changes and process optimization. Traditionally, business processes are specified in natural languages so they cannot be easily validated or optimized; and the decision-making depends more on experience, intuition and artistic talent.

Basic Definition of Analytics

Analytics is decision-making based on data, analysis, and systematic reasoning [1]. Put more concisely: analytics is data-driven decision-making.

Analytics mainly contributes to the meta-process for the creation, adaptation and management of business processes. To some extent, analytics plays the role of sensors and feedback in the modern control theory: based on the sensors and feedback data a missile could automatically pilot through automatic adjustment of its parameters and hit a moving target regardless of various kinds of interferences. The successful deployment of analytics could make a business or organization more agile and achieve efficiency or profit maximization. Analytics is a holistic and hierarchical discipline, stretching from business strategies to operational data collection and integration [2].

Typically, the key performance indicators (KPIs) for processes are identified based on an enterprise value proposition or business objectives (business modeling), and the values of the KPIs are collected and analyzed to support decision-making in business process execution. The scientific foundation of analytics includes statistics, operations research, data mining, and data visualization. Modern IT infrastructure and tools, including Enterprise Resource Planning (ERP), Business Process Modeling (BPM) [4], data warehouses with Online Analytic Processing (OLAP) capabilities [6], and enterprise-scope data analysis and visualization platforms [7], make integrating real-time decision-making based on big or unstructured data into business process execution possible.

Example: A university runs many collaborating processes to implement its value proposition. Enrollment, graduate satisfaction index, completion rate, and return on investment (ROI) are examples possible KPIs. The university strategic plan could be aspirational, operational, data-driven, or a combination of the above. The Schools/Colleges may further identify conversion rate, retention rate, research productivity, community visibility, and the current student satisfaction index as additional KPIs in their implementation plans, and set verifiable targets for all of the KPIs. The IT department maintains a data warehouse of the current and historical data of the KPIs as well as market data like job indices and competitor information. The university analytics platform automatically distills information from the available data and provides the administrators a real-time university performance dashboard so they can promptly identify and solve problems and modify the business processes in response to economic/market changes.

Analytics’ Relationship with Related Methodologies and Technologies

Since the end of World War II, business efficiency and business quality became the priorities of the industrial revolution. Operations research, business process management methodologies, and modern analytics IT platforms are the three major approaches to improve business efficiency and quality.

Operations research, also called management science or decision science, uses mathematical modeling, statistical analysis, mathematical optimization [8] and game theory to optimize decision-making in the project scope. The example applications of operations research include resource assignment, vehicle routing, supply chain management, chip circuit layout, and project management.

Since late 1950s the leading American and Japanese manufacturers have been developing best practices in business process management culminating in the form of Total Quality Management (TQM) [9] and Lean Six Sigma (LSS) [3]. LSS is currently embraced by all Fortune 200 companies, including the service industries, and managed by American Society of Quality (http://asq.org), the global community of experts and the leading authority on quality in all fields, organizations, and industries. LSS extended the scope of decision-making from project-scope optimization to include more strategic and enterprise-scope challenges. Such challenges include data-driven and value proposition driven identification of enterprise priorities, integrating data-driven decision-making in business processes, measurement system analysis, knowledge discovery, statistical process control, process capability analysis, meta-processes for business process life cycle management, and business process management professional training and certification.

Although analytics started as a subset of LSS focusing on data-driven decision making, it has expanded its scope to significantly overlap with the rest of LSS. The advance of information technologies supports the codification of business processes through Business Process Modeling (BPM [4], BPMN [5]) so processes can be validated and optimized. The coded processes can run by a generic business engine and be integrated with the collaborating processes and other information systems. ERP, BPM, and data warehouses and Online Analytical Processing (OLAP) [6] support enterprise-scope analytics, real-time transaction/performance/audit data collection and analysis, and decision-making through rule-based AI (artificial intelligence) engines.

Why Analytics is Important for Pace University Curricula

Pace University has its root in professional business education, and its Lubin School of Business is a brand name for education excellence. Pace is positioned to produce future business/industry leaders, and its graduates are expected to take on more administrative responsibilities as they advance on their career paths. The introduction of analytics in the Pace curricula in the form of elective courses, minors, or majors would make Pace graduates more business and administration savvy, and thus would be better positioned for enterprise innovations.

The following are sample skills that could be made available to the students by introducing analytics into our programs:

For Lubin students: credit scoring, fraud detection, pricing, supply chain optimization, demand forecasting, customer profitability, and performance management.

For Dyson students: new product development, experiment design, non-profit organization management, and performance management.

For Seidenberg students: enterprise analytics platforms, data warehousing, automatic process execution and performance data collection, web analytics, and performance management.

For Education students: admission management, retention management, education budget/resource allocation, course quality management, effective online education, and performance management.

For College of Health Professions students: drug interaction, preliminary diagnoses, disease management, risk management, customer relationship management, and performance management.

Proposed Strategy for Developing Pace University Analytics Programs

Analytics is a highly inter-disciplinary domain and calls for faculty expertise from all Pace Schools and Colleges. It also has the potential to enhance the program quality in all Pace Schools and Colleges. Different people or companies may have different perspectives on analytics, and the importance of analytics is due to its effectiveness in addressing the real world challenges for enterprise innovation.

The following are the initial steps that we plan to focus:

  • An e-platform for all Pace faculty to participate in brainstorming as well as knowledge sharing and development. In the coming months the core participating faculty with passion will be identified to form a steering committee for this project.
  • The project should start with forging consensus among the faculty on the main body of knowledge and the way to move this initiative forward. With this article as a primer, we can start a brainstorming process where the faculty general is welcome to add their opinions and comment on the opinions of their peers. A discussion forum has been created here for this discussion.
  • We also need to identify all Pace expertise by inviting our faculty to post descriptions of courses or course modules that they are interested in developing or sharing. Courseware sharing across the Schools/Colleges is important so we don’t reinvent the wheel. Courseware sharing at the course module level may help solve the course credit limit problem. A dedicated discussion forum has been created here for this purpose.
  • We will run a series of seminars and workshops by internal and external domain experts to share the core knowledge identified through this process. IBM has agreed to contribute its analytics briefings and workshops starting from March 2012.
  • Each School/College is encouraged to develop its own strategy for integrating analytics into its programs, but it would be highly desirable to share and reuse available courses or courseware within the university. Seidenberg plans to launch a BS-IT analytics concentration in January 2013.

To make our analytics programs current and unique, it is important to go beyond the traditional spreadsheet-based statistical analysis or project optimization. We need to embrace the big view of business process management and the latest challenges in analytics, and the latter include analytics with big data or unstructured data, real-time analytics integrated in business process execution, and enterprise-scope analytics based on unified data warehouses and analytics frameworks.



  1. Thomas H. Davenport, Jeanne G. Harris and Robert Morison. Analytics at Work: Smarter Decisions Better Results, Harvard Business School Publishing, 2010, ISBN 978-1-422-17769-3
  2. Cert H.N. Laursen and Jesper Thorlund. Business Analytics for Managers: Taking Business Intelligence Beyond Reporting, John Wiley & Sons, Inc., 2010, ISBN 978-0-470-89061-5
  3. Thomas Pyzdek. The Six Sigma Handbook, McGraw-Hill 2003, ISBN 0-07-141015-5
  4. Michael Havey. Essential Business Process Modeling, O’Reilly Media 2005, ISBN 978-0596008437
  5. Thomas Allweyer. BPMN 2.0, BoD 2010, ISBN 978-3839149850
  6. Oracle. Data Warehousing Concepts, http://docs.oracle.com/cd/B10501_01/server.920/a96520/concept.htm, valid by 2/24/2012
  7. IBM. Cognos Business Intelligence and Financial Performance Management, http://www-01.ibm.com/software/analytics/cognos/, valid by 2/24/2012
  8. Lixin Tao. “Research Incubator: Combinatorial Optimization,” CSIS Technical Report 198, Pace University, 2004, http://csis.pace.edu/lixin/dps/combinatorialOptimizationTechReport198.pdf
  9. Christian Madu, handbook of Total Quality Management, Springer 1999, ISBN 978-0-412-75360-2