1. Business Intelligence and How to Teach It
Hugh J. Watson
Terry College of Business
University of Georgia
Parts of this presentation were first given on February 13, 2012 at the TDWI BI Executive Summit in Las Vegas. The complete presentation was given as a keynote talk at the Southern Association for Information Systems conference on March 28, 2012 in Atlanta. Feel free to use any of these slides in teaching your classes.

2. Topics l Terminology, frameworks, and concepts What’s new in BI
Different BI “targets”
Exemplars of BI-based organizations
Requirements for being successful with BI and analytics
What I teach in my BI courses
Using the Teradata University Network to teach BI l

3. What Is Business Intelligence?
Its roots go back to the late 1960s
In the 1970s, there were decision support systems (DSS)
In the 1980s, there were EIS, OLAP, GIS, and more
Data warehousing and dashboards/scorecards became popular in the 1990s
An excellent history of DSS is available at DSS resources. See Power, D.J., “A Brief History of Decision Support Systems,” DSSResources.com, World Wide Web, URL DSSResources.com/history/dsshistory2.8.html, version 2.8, May 31, 2003.
In the 1960s, the first computer applications were for scientific purposes and transaction processing. Reports that summarized the processed transaction data provided some, but very limited, information for decision support.
It was not long, however, before maturing technology, business need, and the vision of the early pioneers led to the first decision support applications. Much of the early development work was conducted at MIT and Harvard. Particularly important was Michael S. Scott Morton’s 1967 doctoral dissertation research. He built, implemented, and tested a system to support planning for laundry equipment. Later, he published an influential Sloan Management Review article and book that helped spread decision support concepts and provided a name for these analytical applications -- management decision systems – though this name was soon to be replaced with decision support systems.
Throughout the late 1960s and early-to-mid 1970s, a variety of decision support applications were developed and reported. Many academicians with backgrounds in management science/operations research were attracted to the field because of the practical potential for these new kinds of applications. Decision support systems (DSS) began to be used to describe these applications and the name of the emerging field. During this time, Ralph Sprague published his initial DSS frameworks articles that were followed by his seminal MIS Quarterly article and book.
Over the next decade, other kinds of decision support applications emerged – executive information systems, online analytical processing, and geographic information systems to mention a few.
A major stumbling block to supporting decision making was the lack of a decision support data infrastructure. This was addressed with the creation of data marts and warehouses in the 1990s. Also in the 1990s, dashboards and scorecards appeared, and in some ways (e.g., the providing of performance metrics), replaced EIS.

4. What Is Business Intelligence?
Howard Dresner, a Gartner analyst, coined the BI term in the early 1990s
Today there is much discussion of analytics
There are many BI definitions, but the following is useful
Though the roots of the business intelligence term can be traced earlier, Howard Dresner is commonly given credit for it when he was an analyst at Gartner and used it in his writings.
Today, the analytics or business analytics term is “hot.” It is discussed in detail later.
There is no commonly agreed definition for business intelligence, but the following is representative and useful.

5. Business intelligence (BI) is a broad category of applications, technologies, and processes for gathering, storing, accessing, and analyzing data to help business users make better decisions.
This definition is broad. BI encompasses not only applications, but also technologies and processes. It includes not only “getting data out” (through tools and applications), but also getting “data in” (to a data mart or warehouse).
It should be pointed out that some authors use the BI term to refer to “getting data out” and data warehousing as “getting data in.” And there are some authors who use the data warehousing term to refer to both “getting data in” and “getting data out,” much like we are using the BI term. The good news is that the differing terminology does not normally cause confusion because of the context in which the terms are used.

6. Different BI targets require different BI environments
Different BI targets require different BI environments. For example, developing a single or a few BI applications may require only a data mart rather than a data warehouse. This slide shows a generic BI environment. At the left are the source systems that provide data to the decision support data repository (i.e., data warehouse and marts). Data integration technology and processes are needed to prepare the data for decision support use. The mart or warehouse can employ a variety of architectures, technologies, and data models. On the right, a variety of users can access the data using different tools and applications. To ensure that BI meets its intended purposes, metadata, data quality, and governance processes must be in place.

7. Things Are Getting More Complex
Source systems include social media, machine sensing, and clickstream data (Big Data)
The cloud, Hadoop/Reduce, and appliances are being used as data stores
Advanced analytics are growing in popularity and importance
Current BI architectures are not your “mother’s architecture” as of recent as five years ago. In many organizations they are much more varied and complex.
Organizations are finding business value in capturing, storing, and analyzing new kinds of data, such as social media, machine sensing, and clickstream. Because of its three Vs -- volume, variety, and velocity – this kind of data is often called Big Data. Many companies are performing new kinds of analytics, such as sentiment analysis to better and more quickly understand and respond to what customers are saying about them and their products.
There are changes in how this data is stored and analyzed. For example, some companies store and analyze data in the cloud. The open source software Hadoop and MapReduce from Apache Software is being used to store and analyze massive amounts of multistructured data (as opposed to the more structured data maintained in RDMS).

8. Advanced Data Visualization BI Based Organizations BI Governance
BI in the Cloud
Big Data
Data Appliances
Pervasive BI
Columnar Databases
Mobile BI
Predictive Analytics
Real Time BI
Advanced Data Visualization
BI Based Organizations
BI Governance
In-Memory Analytics
Data Scientists
Rules Engines
SaaS
BI Competency Centers
Hadoop/MapReduce
There is much new in BI. The technologies, applications, people, and processes shown in this slide are some of the most interesting.
BI 2.0 Software
BI Search
Agile
Open Source BI Software
Text Analytics
Master Data Management
Event Analytics

9. What Is Meant by Analytics?
A new term for BI
Just the data analysis part of BI
“Rocket science” algorithms
Three kinds of analytics
The analytics or business analytics term can be interpreted in different ways. The distinctions become most important when discussing what is required for being successful with advanced analytics (predictive and prescriptive analytics, which are discussed later).

10. Descriptive Analytics
Descriptive analytics, such as reporting/OLAP, dashboards, and data visualization, have been widely used for some time. They are the core of traditional BI.
What has occurred?

11. Predictive Analytics What will occur?
Algorithms for predictive analytics, such as regression analysis, machine learning, and neural networks, have also been around for some time. Recently, however, software products have made them much easier to understand and use. They have also been integrated into specific applications, such as for campaign management. Marketing is the target for many predictive analytics applications. Descriptive analytics, such as data visualization, is important in helping users interpret the output from predictive and predictive analytics. Prescriptive analytics are often referred to as advanced analytics.
What will occur?

12. Prescriptive Analytics
Prescriptive analytics provide an optimal solution, often for the allocation of scarce resources. They, too, have been in academia for a long time but are now finding wider use in practice. For example, the use of mathematical programming for revenue management is common for organizations that have “perishable” goods (e.g., rental cars, hotel rooms, airline seats). Harrah’s has been using revenue management for hotel room pricing for some time.
What should occur?

13. There are different “targets” for BI
Companies can have different “targets” or organizational approaches to analytics. Each target can deliver value and may be appropriate for a particular company.

14. A single or a few applications
A point solution
May be departmental
Serves a specific business need
A possible entry point
This is a common starting point for analytics in an organization. It can serve as a proof of concept for analytics. A business unit has a specific need and puts the resources together to create an analytics solution. Typically, this is a point solution for that need. Over time, there are more point solutions and management becomes aware of the need to put resources, structures, and processes around analytics. This can lead to enterprise-wide analytic capabilities.
A 2011 Bloomberg BusinessWeek study found that analytics is still in the emerging state in organizations.

15. Enterprise analytical capabilities
The infrastructure is created for enterprise-wide analytics
Analytics are used throughout the organization
Analytics are key to business success
With this target, a company recognizes the need to treat analytics as an enterprise resource in order to compete in the marketplace an puts and analytics infrastructure in place – data, tools, technology, people, governance, processes, etc.

16. Organizational transformation
Brought about by opportunity or necessity
The firm adopts a new business model enabled by analytics
Analytics are a competitive requirement
This is the most interesting target. It occurs when a company either sees an opportunity or faces a problem that requires a new business model that can only be executed using analytics. To an extent, analytics is the business. Without analytics, the business model cannot be execute and the company cannot compete.

17. For BI-based organizations, the use of BI/analytics is a requirement for successfully competing in the marketplace.
In some industries, the use of advanced analytics has moved beyond a “nice to have” to being a requirement.

18. 2011 Academic Research 5-6% Firms that emphasize Productivity data and
analytics
5-6%
Productivity
Return on equity
Market value
The relationship between the use of data and analytics in decision making and a variety of organizational performance measures is described in a 2011 study by Brynjolfsson, Hitt, and Kim in the Social Science Research Network (SSRN).
Also, A 2010 IBM/MIT Sloan Management Review research study found that top performing companies in their industry are much more likely to use analytics rather than intuition across the widest range of possible decisions.
A 2011 TDWI report on Big Data Analytics found that 85% of respondents indicated that their firms would be using advanced analytics within three years.
A 2011 IDC study found that analytics is one of the top two IT priorities for this year.

19. Conditions that Lead to Analytics-based Organizations
The nature of the industry
Seizing an opportunity
Responding to a problem
There are several reasons that some companies are heavily committed to analytics – the nature of the industry in which they compete, an opportunity, or a critical business problem.
Tom Davenport in his books, Competing on Analytics (with Jeanne G. Harris) and Analytics at Work (with Jeanne G. Harris and Robert Morison), has shown how a variety of companies are benefiting from analytics.

20. Complex Systems versus Volume Operations
A distinction made by
Geoffrey Moore
Helps in understanding
what kinds of organizations
are most likely to be
analytics based
There are reasons why companies in some industries rely more on analytics to compete than companies in other industries. Geoffrey Moore in his books Crossing the Chasm and Inside the Tornado makes a distinction between companies that operate complex systems versus volume operations.

21. Complex Systems
Tackle complex problems and provide individualized solutions
Products and services are organized around the needs of individual customers
Dollar value of interactions with each customer is high
There is considerable interaction with each customer
Examples: IBM, World Bank, Halliburton
Companies that operate complex systems use analytics, but they are less reliant of analytics to achieve competitive advantage.

22. Volume Operations
Serves high-volume markets through standardized products and services
Each customer interaction has a low dollar value
Customer interactions are generally conducted through technology rather than person-to-person
Are likely to be analytics-based
Examples: Amazon.com, eBay, Hertz
Analytics are critical to high volume companies competing in the marketplace.

23. The nature of the industry: Online Retailers
BI Applications
Analysis of clickstream data
Customer profitability analysis
Customer segmentation analysis
Product recommendations
Campaign management
Pricing
Forecasting
Dashboards
Online retailers like Amazon.com and Overstock.com are great examples of high volume operations who rely on analytics to compete. As soon as you enter, their sites a cookie is placed on your PC and all clicks are recorded. Based on your clicks and any search terms, recommendation engines decide what products to display. After you purchase an item, they have additional information that is used in marketing campaigns. Customer segmentation analysis is used in deciding what promotions to send you. How profitable you are influences how the customer care center treats you. A pricing team helps set prices and decides what prices are needed to clear out merchandise. Forecasting models are used to decide how many items to order for inventory. Dashboards monitor all aspects of organizational performance.

24. “We are a business intelligence company”
Patrick Byrne,
CEO, Overstock.com
When I interviewed Patrick Byrne and asked him to describe Overstock.com, I expected him to say it was an “online retailer” or perhaps an “information-based company”. Instead, he said that it is a “business intelligence company”. That is how Overstock.com competes in the marketplace, as does other large online retailers.

25. Seizing an Opportunity: Harrah’s
In 1993, the gaming laws changed
Harrah’s decided to compete and expand using a brand and customer loyalty strategy
Implemented WINet with an ODS and DW
Offered the industry’s first customer loyalty program, Total Rewards
Harrah’s was originally the “blue collar” casino where everyone knew your name. When the gaming laws changed, Harrah’s adopted a new business model that included expansion, creating a brand identity, and the industry’s first loyalty program, Total Rewards. Gary Loveman, a Harvard marketing professor, was brought in to drive this new business strategy. Key to this new model was analytics using an operational store and a data warehouse. It allowed Harrah’s to perform analytics in order to know who its customers are, where they gamble, what games they play, their profitability, and what offers to make to get them to return.

26. Seizing an Opportunity: Harrah’s
Fact based decision making replaced “Harrahisms”
Today it is the largest gaming company in the world
Recently renamed Caesars
The managers of the Harrah’s properties used to run their casinos as private fiefdoms and decisions were made based on “Harrahisms” – things that were just assumed to be true. With the new business model, decision making became much more centralized and was based on constant experimentation of what worked best (i.e., fact based decision making). The success of this approach is seen in Harrah’s now being the largest gaming company in the world. The “blue collar” casino bought the more upscale Caesars in 2004 and changed its corporate name to Caesars in Quite a success story. Many of the people who were successful with Harrah’s original uses of analytics were hired by other casinos and spread the use of analytics throughout the gaming industry.

27. Responding to a problem: First American Corporation
The bank was failing
A new management team stopped the bleeding
A customer intimacy strategy was implemented, Tailored Client Solutions
First American was a regional bank headquartered in Nashville, TN. Don’t look for it now, it has been bought and renamed several times. The bank had lost $60M and was operating under letters of agreement with regulators. It had serious problems. A new management team was brought in that cut costs and stopped the bleeding. Management knew, however, that it needed a new business strategy if the bank was to survive. As it explored its options (e.g., being a low cost provider), it realized that its best bet was a customer intimacy strategy (supported by analytics) because of the information it had about its customers that other banks didn’t. This strategy was called Tailored Client Solutions.

28. Responding to a problem: First American Corporation
The business strategy was enable by a data warehouse and BI
A data warehouse and extensive analytics were required to implement the new business strategy.

29. Responding to a problem: First American Corporation
External talent was brought in as needed
Applications using VISION were developed for every component of TCS
The bank was transformed from “banking by intuition” to “banking by information and analysis”
First
American
A heavy dose of consultants and professional services were brought in to quickly build the warehouse and do the customer and product profitability analyses. Small teams also worked on developing applications in each of the TCS application areas. Each of the applications provided financial “lift” and gave credence to this new approach to running the bank. Decision making became fact based. The bank moved from financial disaster to being a leader in the financial services industry and won the prestigious Society for Management Information competition for its work.

30. Let’s Answer Two Questions
What is special about advanced analytics?
What are the requirements for being a BI or analytics-based organization?
The factors for success with analytics, such as executive support and sponsorship, are largely the same as for BI. The differences are in the details. Yes, the devil is in the details.
The following list of factors is inspired by an article written by Steve Williams, “Assessing BI Readiness: A Key to BI ROI,” Business Intelligence Journal, Summer 2004.

31. A clear business need
We all know that almost every thing should be a business rather than a technology solution. At Harrah’s, it was enabling its opportunity to expand into new markets, encourage cross play at Harrah’s casinos, and increase customer loyalty and profitability. At First American, it was to “save the bank” through a customer intimacy strategy supported by a data warehouse and analytics. In both of these cases, and in many others where analytics is used, the business need is to focus on customer needs and preferences.
The driver for analytics in many companies is a specific organizational pain. Many companies use analytics in marketing in order to better understand and respond to customer needs and preferences. According to a 2011 Bloomberg BusinessWeek study, analytics are also commonly used in finance and strategy/planning.
I’m convinced that integrating analytics into operational systems in the form of rules will become increasingly important.

32. Strong, committed sponsorship
We also know that if you don’t have strong, committed sponsorship, it is difficult to succeed. If the target is a single or a few analytical applications, the sponsorship can be departmental. However, if the target is enterprise-wide analytics, or especially organizational transformation, sponsorship needs to be at the highest levels and organization wide.
A 2011 B

33. Alignment between the business and IT strategy
At Overstock.com, Harrah’s, and First American, the business and analytics strategies were so intertwined that it is impossible to discuss one without the other. The connection might not always be this close, but it is important to make sure that the analytics work is always supporting the business strategy.

34. A fact-based decision making culture
In a fact-based decision making culture, the numbers rather than intuition drive decision making. There is also a culture of experimentation to see what works and doesn’t. At Overstock.com different web site designs are tested. At Harrah’s different offers to different market segments are tested.

35. Creating a Fact Based Culture
Things that senior management needs to do:
Recognize that some people can’t or won’t adjust
Be a vocal supporter
Stress that outdated methods must be discontinued
Ask to see what analytics went into decisions
Link incentives and compensation to desired behaviors
At First American, prior to Tailored Client Solutions, there were 12 people in Marketing. Afterwards, there were 12, but none the same. As the CEO explained it, “The original group thought that marketing was giving out balloons and suckers along the teller line and running focus groups.” They either left the bank or took other positions and were replaced by people who could do the analytical marketing work.
At Harrah’s, three things will get you fired – sexual harassment, stealing, and not testing your theories.
A 2011 Blumberg BusinessWeek study found that culture plays a critical role in the effective use of analytics. Trust in analytics is a big issue. The study found that just 58% of respondents believe that executive management trusts the results of analytics.

36. A strong data infrastructure
Predictive analytics requires data for initial analysis, model development, model testing, and model maintenance. This data often needs to be raw and detailed. Many companies have a data warehouse in place and its existence facilitates predictive analytics. But providing a strong data infrastructure is getting more complex. BI Directors need to think about the role and use of appliances, the cloud, software as a service, columnar databases, in-memory analytics, in-data base analytics, grid computing, data federation, and more. Some of the advanced analytics are best served by analytical sandboxes that allow data to be analyzed on the desktop or in physical or logical data repositories. Then there is the need to accommodate Big Data and new technologies like Hadoop and MapReduce. The picture is of Yahoo’s Hadoop cluster. The International Institute for Analytics predicts that Big Data analytics will top all other areas of growth in analytics during 2012 due to the rapid expansion of social, mobile, location and transaction-based data. There is a shortage of people who have the skills for working with and analyzing Big Data.
While we normally think that predictive analytics and data warehouses go together, this isn’t always the case. Scientific researchers have developed robust predictive models for years using very specific data sets. This use is a bit of an outlier, however.
A 2011 Bloomberg BusinessWeek study found that data is the number one challenge in the adoption of analytics.

37. This slide comes from Wayne Eckerson’s (2011) excellent research report on Big Data analytics. It depicts Eckerson’s thinking about what future analytical architectures will look like. The top of the slide shows casual users accessing descriptive analytics based on warehouse data and also streaming data that is analyzed and displayed on dashboards or results in alerts. At the bottom, power users, such business analysts and data scientists, access a wide variety of data sources, including Hydoop/Map Reduce and analytical sandboxes.
Source: Eckerson, 2011

38. The right analytical tools
Most organizations’ BI environments are designed to support descriptive BI. As organizations go into predictive analytics and expand the types of data they analyze, it is necessary to add analysis tools and technologies to their analytics’ architecture. Traditional BI tools and predictive analytics are very synergistic, however. BI tools are very useful in understanding the data and thinking about relationships before using predictive analytics. Data visualization tools are very useful in helping interpret the output for predictive analytics models.
While you may want to standardize on a single of a few advanced analytics tools, modelers often have preferences for specific tools they are familiar with and are well suited for a specific task. They also like to experiment with new ones, often from the open-source community. It is a balancing act between the benefits of standardization and getting the best tool for the task at hand.

39. New tools and architectures may be needed
While traditional BI vendors claim their tools support data mining/predictive analytics, this is often not true. Slicing and dicing and data visualization are not data mining. Data mining requires tools that incorporate algorithms and processes that are designed specifically to find hidden relationships in data. SAS and SPSS are two of the traditional leaders in this space. R is a programming language and software environment for statistical computing and graphics and is now the most popular tool used by data miners. It is also at the core of many open source products. As organizations expand their analytics to include multi-structured data for applications such as sentiment analysis, organizations are adding to their architectures with new advances such as Hadoop/MapReduce. They also are turning to analytical sandboxes, in-memory analytics, and analytical appliances to provide an analytics environment and power that data scientists need. It should also be remembered that exploring, preparing, and combining data is often required prior to running a predictive analytics algorithm and the tools should support this requirement. The ability to score based on the results of predictive modeling is important for many applications, such as customer credit.

40. Strong analytical personnel in an appropriate organizational structure
The use of advanced analytics will require new skills. A 2011 Bloomberg BusinessWeek study found that many organizations lack the proper analytical talent. With proper training some of the existing personnel will be able to accept the challenge, at least for more structured analytics supported by appropriate software. Most of the large vendors, such as IBM and SAS, offer training on their products. For the “rocket science” work, new personnel will often have to be brought in, either through hires or professional services. This was the case at both Harrah’s and First American.
Another option is to outsource your analytics to providers such as Mu Sigma and Apollo Data Technologies. Companies that take this approach don’t have to invest time and money in hiring, training, and organizing their own analytic teams. On the other hand, companies may not feel comfortable turning their data to a third party and it can get expensive if there is a lot of analytics work to be done.

41. Knowledge Requirements for Advanced Analytics
Business Domain
Analytics requires an understanding of what data is available, how to access it, and how to manipulate it. Data is necessary for first building models and later testing them. Choosing the right data to include in models is important. Predictive analytics should not be searching for a diamond in a coal mine. You will get too many spurious findings. Rather, it is important have some thoughts as to what variables might be related. And once you have findings, domain knowledge is necessary to understand how they can be used. Consider the hoary story of the relationship between beer and diapers in the market basket of young males in convenience stores. You still have to decide (or experiment to discover) whether it is better to put them together or spread them across the store (in the hope that other things will be bought while walking the isles). Domain knowledge is important here. And finally, it is necessary to understand the models being used. At a minimum, this requires training in multivariate statistics.
Data
Modeling

42. Business Analyst
Uses BI tools and applications to understand business conditions and drive business processes
Because of the need for predictive analytics, companies want their business analysts to move more into this application area. Vendors have responded to this demand in various ways. They are providing more user-friendly graphical interfaces and automating the process of building models. Some products build multiple models using different algorithms and suggest which one is best based on the data. Training is critical to upgrade the skills of the business analysts. Other vendors are offering predictive analytics using the software as a service model. While these advances help, for the work that requires a deep understanding of statistics, a data scientist is needed.
Big Data and Hadoop/MapReduce will also create a demand for people with specialized skills.

43. Data Scientist
Uses advanced algorithms and interactive exploration tools to uncover non-obvious patterns in data
The data scientist title is taking hold, even though it sounds elitist . Data scientists have advanced training in multivariate statistics, artificial intelligence, machine learning, mathematical programming, and simulation to perform predictive and prescriptive analytics. They often hold advanced degrees, including PhDs in econometrics, statistics, mathematics, and management science. You don’t need a lot of them, but for some of the really advanced work, they come in very handy. Be prepared to pay top dollar for them, though. A number of universities are ramping up to meet the demand, such as the new Master of Science in Analytics at North Carolina State University.

44. Business Domain Modeling Data Data Business Analyst Data Scientist
While business domain, data, and modeling skills are needed by business analysts and data scientists alike, their relative importance changes. For example, business analysts must have a high level of business domain knowledge while data scientists must be high in modeling skills. To make sure that all skills are present on a project, it is wise to have both business analysts and data scientists working on it.
Data scientist will also need permissions to access a wider range and volume of data than for traditional BI applications. They will need access to detailed data that can be aggregated, joined, or transformed in the ways needed for their work.
Modeling
Data
Data
Modeling

45. Education: BBA, MBA MS, PhD
Business Analyst
Data Scientist
Education: BBA, MBA MS, PhD
Tools: Cognos, Hyperion KXEN, SAS
Analytics: OLAP Neural networks
Focus: Business Analytics
Data scientists work on building products (e.g., predictive models put into production) and finding interesting insights in the data. Business analysts focus on measuring and managing business performance.
The backgrounds, interests, and career paths for business analysts and data scientists tend to be different. Most fundamentally, the business analysts focus is on the business and see their career path as moving through a business unit. The data scientists focus on their technical expertise and see their paths as working on a variety of analytics projects across the organization.
Some companies have had good luck having MBA-educated marketing experts work hand-in-hand with statistical modelers. Business knowledge tends to rub off on the modelers and the MBAs gradually pick up technical skills.
Scope: Departmental Enterprise-wide
Value: High Exceptionally high

46. Where to put the analytics team?
Spread throughout the organization
In a standalone unit
In some form of an Analytics Competency Center
The most common approach is to spread the team(s) throughout the organization, putting the people where the need for analytics is the greatest. For example, a team focusing on customer churn and other marketing-related analyses might be part of the marketing department. A concern is that the resources as a whole are not being used in an optimal manner. If there are multiple teams in an organization, it is important that they communicate and share best practices, expertise, models, etc.
At Overstock.com, all of their BI specialists are spread throughout the business and work directly with their business counterparts. This approach is consistent with agile development concepts.

47. What I Teach in My BI Course
Concepts, terms, and definitions
Making the business case for BI
Development methodology for BI
Data and data warehousing
BI software
Interface design
BI applications (e.g., dashboards)
Analytics
Best practices case studies
Organizational issues
Determining the ROI for BI
Implementing BI enterprise wide
Future directions for BI
I currently teach a 3 credit hour undergraduate BI course for MIS majors and a 1.5 credit hour course for MBA students. Some of the materials in the two courses are the same; for example, some of the BI software used is the same. However, because the undergraduate course is for more credit hours and for MIS majors, it takes more of a “providers” rather than a “users” perspective, is a little more technical, and is more detailed.

48. The Teradata University is sponsored by Teradata, a leader in the data warehousing and BI marketplace.

49. Teradata University Network
A premier, free online educational resource for university professors around the world who teach classes on data warehousing, DSS/business intelligence, and database.
Current Membership
Over 3,000 registered faculty members
Representing 1,641 universities
In 90 countries
Thousands of students
An international community, led by academics, whose members share their ideas, experiences, and resources with others
The Teradata University Network (TUN) is a free portal for faculty and students in BI, data warehousing, and database courses. Unlike other vendor-sponsored programs/alliances, TUN is led by leading academics. To register to use TUN, go to

50. Using the Teradata University Network
Faculty apply for membership, and are authenticated
Faculty have access to course syllabi, articles, cases, projects, assignments, presentations, software (Teradata, MicroStrategy) various datasets, web seminars, and more.
Faculty have the ability to post and share their favorite content
Faculty send students to TUN to access course-related materials
Both faculty and students must register to use TUN. Faculty are authenticated to ensure they are faculty and should be granted access to resources only for faculty. TUN has both faculty and a student “views,” and while the views share many of the same resources, the faculty view includes resources only for faculty, such as teaching notes and solutions to cases and assignments.
The resources on TUN are contributed by Teradata, other partners, TUN academic Board members, and faculty around the world. It is a highly collaborative resource. You are encouraged to share your favorite course content on TUN.
In general, faculty use TUN to find resources for their courses, and then send students to TUN to use the resources.

51. Resources from TUN Articles Current state of BI Business analytics
Big data
Future directions for BI software
Understanding users value proposition
Decision support sweet spot
Dashboards and scorecards
Dashboard design
Data warehousing
Data profiling
Data quality
Data mining primer
Assessing BI readiness
Business schools need to
change what they teach
I don’t use a textbook in my BI courses. About 90 percent of the resources come from TUN. You can see my course syllabi on TUN. This slide contains the articles on TUN (not the exact titles but close) that I use.
Many of the BI texts have some degree of integration with TUN, such as the inclusion of assignments from TUN.

52. Resources from TUN Harrah’s First American Corporation
Cases
Harrah’s
First American Corporation
Continental Airlines
Retailstore.com
Catalina Marketing
Norfolk Southern Railway
Spokane Teachers Credit Union
U.S. Xpress
Videos
Applebee’s
Nationwide
BSI: Retail Tweeters
The cases on TUN tend to be illustrations of best practices. Some are long, such as Harrah’s, while others are described in articles, such as about U.S.Xpress. This presentation includes materials taken from the Harrah’s, First American Corporation, and RetailStore.com cases.

53. Resources from TUN 1-800 CONTACTS Genericorp Software
Assignments
1-800 CONTACTS
Genericorp
Software
Teradata SQL Web Assistant
MicroStrategy
Tableau
I use Teradata SQL Web Assistant for a quick review of relational data bases and SQL. MicroStrategy is used to illustrate a leading reporting/OLAP tool, and Tableau for data visualization. All of these are “packaged” for easy use in your classes.

55. References
Brynjolfsson, Hitt, and Kim, “Strength in Numbers: How does data-driven decision-making affect firm performance?,” Social Science Research Network (SSRN), April 2011.
Cooper, Watson, Wixom, and Goodhue, "Data Warehousing Supports Corporate Strategy at First American Corporation," MIS Quarterly, December 2000.
Eckerson, “Big Data Analytics,” BeyeNetwork, September 2011.
Davenport, Harris, and Morison, Analytics at Work: Smarter Decisions, Better Results, Harvard Business School Press, 2010.
These are references that support the presentation and provide more detail.

56. References
Davenport and Harris, Competing on Analytics: The New Science of Winning , Harvard Business School Press, 2007.
Eckerson, W. (2011). Big Data Analytics: Profiling the Use of Analytical Platforms in User Organizations. BeyeNetwork.
LaValle, et al., “Analytics: The New Path to Value,” IBM, MIT Sloan Management Review, 2010,

57. References
Moore, Crossing the Chasm: Marketing and Selling High-tech Products to Mainstream Customers, HarperBusiness Essentials, 2002.
Moore, Inside the Tornado, HarperBusiness Essentials, 2004.
Watson, “Business Analytics Insight: Hype or Here to Stay?” Business Intelligence Journal,” March 2011.
Watson and Volonino, “Harrah’s High Payoff from Customer Information,” Printed in Eckerson and Watson, Harnessing Customer Information for Strategic Advantage: Technical Challenges and Business Solutions, TDWI, 2000.

58. References
White paper, “The Current State of Business Analytics: Where Do We Go From Here?” Bloomberg BusinessWeek Research Services, 2011.
Williams, “Assessing BI Readiness: A Key to BI ROI,” Business Intelligence Journal, Summer 2004.

59. Dr. Hugh J. Watson is a Professor of MIS and a holder of a C
Dr. Hugh J. Watson is a Professor of MIS and a holder of a C. Herman and Mary Virginia Terry Chair of Business Administration in the Terry College of Business at the University of Georgia. Hugh has authored 23 books and over 150 scholarly journal articles. He is a Fellow of the Association for Information Systems and The Data Warehousing Institute and is the Senior Editor of the Business Intelligence Journal. For the past 20 years, Hugh has been the consulting editor for John Wiley & Sons’ MIS series. He can be contacted at
Feel free to contact me if you have any questions about BI or the Teradata University Network.