White Paper

Moving Past Traditional BI to Self-Service AnalyticsHigher education harnesses the benefits of instantaneous decision making

ContentsWhat’s Wrong With ‘Just BI’? ..........................................1

Barriers to Analytics: IT Perspective ..............................1

Barriers to Analytics: End-User Perspective ................2

Moving Past Traditional BI to Self-Service Analytics ........................................................2

Self-Service Analytics in Action .....................................3

How SAS® Can Help .........................................................3

Case in Point: Western Kentucky University Gains Insight Into Its Overall Performance ..................5

Learn More ........................................................................5

1

The image of the staid, ivory tower is fast receding in higher education as schools learn to adapt quickly to both internal and external pressures. With greater accountability and fiercer competition for students, historical reporting is no longer suffi-cient to anticipate and respond to the rapid changes in educa-tional trends and student populations.

Colleges and universities collect vast amounts and types of data – but it’s spread across different schools, departments and campuses in various formats and a multitude of systems. It’s a massive challenge to make all that data available, connected and usable to everyone who needs it. Traditionally, IT (and institutional research) departments stretched their resources as best they could to support reporting requests, but this took time, rendering the information less useful. It is no longer viable for would-be users to wait for IT to free up resources to deliver results. Higher education institutions want to evolve into using their data for a more proactive and interactive flow of informa-tion for decision making at all levels – at a faster pace.

Business intelligence (BI) and analytics tools can provide critical support to help colleges and universities give users much-needed self-reporting access to data and information for stra-tegic decision making.

This paper examines the barriers to adoption from an IT and end-user perspective, and shows how self-service analytics in general – and SAS Visual Analytics in particular – can eliminate these barriers. Self-service analytics empowers users to truly exploit the wealth of data available to them, while ensuring that the IT organization maintains governance and control over that data.

What’s Wrong With ‘Just BI’? Since the term “analytics” can have multiple meanings, a brief clarification is in order. In this paper, analytics refers to a data-centric approach to decision support that supplements more conventional BI. By and large, BI looks backward, telling users what has happened based on historical data contained in data warehouses, which is then presented in predetermined reports. In contrast, analytics looks forward, telling users what is statistically likely to happen in the future and taking into account data that changes daily to explore what-if scenarios, predict outcomes, and solve specific business problems in an interactive, iterative manner. The following chart summarizes the key differences:

BI Analytics

Purpose Information gathering

Problem solving

Orientation Historical, backward-looking

Predictive, forward-looking

Data Aggregated Detailed

Presentation Passive Interactive, iterative

Table 1. The differences between BI and analytics – historical vs. predictive.

Barriers to Analytics: IT Perspective Seemingly, a self-service decision support system that gives managers access to ever-changing data so they could make better decisions would be a proverbial no-brainer. However, from the IT department’s perspective, there are some perceived barriers that stem from the historical analytics approach:

• Complexity. For many IT professionals, the term “analytics” is synonymous with OLAP cubes. While valuable, these cubes can present problems to a busy IT or institutional research (IR) department. The cubes are often complex, which means that their creation can be time-consuming. Once created, there’s no guarantee that they’ll meet the needs of the intended end user. This can happen because of miscommunication between IT and the requesting depart-ment, or because of the fundamentally iterative nature of analytics. More often than not, one answer gives rise to a new question, which in turn means more work for IT. Also, given the complexity of the tools used to create OLAP cubes, many department heads or deans conclude that they simply don’t have the time – much less the skills – to attempt analytics.

• Multiple disparate data sources. Typically, conventional BI approaches to creating reports that involve multiple data-bases require that the data goes through an ETL process. Manually converting that data for compatibility in real or near-real time can be an impossible task using conventional means.

• Performance. Many institutions already produce standard reports that require hours of computer time. On the surface, it would seem that analytics would be even more demanding and may lock up end users’ computers for hours while performing complex calculations.

2

• Analytic talent. The demand for analytical talent is projected to outpace supply by 60 percent in the next few years. Competing with the private sector, colleges and universities are challenged to hire and keep master’s- and PhD-level graduates who can support the extensive data manage-ment and statistical programming needed to meet reporting demands for all departments across the institution.

• Mobile requirements. To gain acceptance among end users in today’s world, web and mobile capabilities are a require-ment, and these capabilities add yet another layer of diffi-culty to implementing BI and analytics. Furthermore, users expect more than simple access to data wherever they may be; they also expect to interact with it.

With self-service analytics, all these challenges are alleviated for IT. With reliable data integration and management, dynamic self-service reporting, and data visualization providing the ability to analyze data and drill down in real time, end users are relieving the burden IT once faced to meet institutional analyt-ical needs.

Barriers to Analytics: End-User Perspective Department heads and other end users have also had their own set of challenges in accessing analytics:

• The learning curve. The idea that analytic functions can be shifted to actual users is difficult to accept. Most education administrative professionals have little or no experience with statistical models and predictive algorithms. Many are daunted by the prospect of being expected to engage in complex mathematics to facilitate fact-based decision making. But in reality, this isn’t the case with self-service analytics. These users can access the tools with genuinely simple-to-use outputs that are clear and easy to grasp.

• Lack of timely results. A fact-based approach to solving business problems is typically iterative. If every analytical iter-ation takes hours – as was the case with most BI and analytics solutions – the process of arriving at a useful answer took too long, was too frustrating, and may not have produced results in time. These barriers only added to the chasm between IT and end users.

And again, these long-held beliefs of access to useful analytics, although previously justified, are no longer a factor when dynamic and self-service analytics – as well as data visualization – is available to the decision makers.

The specific goal of self-service analytics is to put the power of analytics in the hands of actual decision makers with a technology that allows them to rapidly explore problems and solutions. Self-service analytics also requires no advanced statistical training, no analytical specialists and only minimal IT involvement.

Moving Past Traditional BI to Self-Service AnalyticsThe days of ask-and-wait-for reports and OLAP cubes for traditional BI are fast becoming a thing of the past. Self-service analytics closes the gap to put the data in the hands of those who need it, when they need it, to enable far more timely data-driven decisions.

First, moving to a self-service analytics framework eliminates the back-and-forth conversation between the user and the IT department that is the cause of so much misunderstanding – and overhead. To be effective, users need answers fast enough so that they can efficiently explore a problem without having to wait hours and hours for each iteration.

And secondly, the greatest benefit is found in allowing many users to explore all relevant data quickly and easily. They should be able to slice and dice the data, look at more options, uncover hidden opportunities, identify key relationships and make more precise decisions faster than ever before. Self-service analytics, ad hoc visual data discovery, exploration and visualization put lightning-fast insights within everyone’s reach.

Traditional BI is limited in that it’s a reactive approach to problem solving. By shifting your institution’s approach so that sophisti-cated analytics are readily available to business users, you are empowering proactive decision making as shown in Figure 1.

REACTIVE

ALERTS

OLAP

AD HOC REPORTS

STANDARD REPORTS

PROACTIVE

OPTIMIZATION

PREDICTIVE MODELING

FORECASTING

STATISTICAL ANALYSIS

Figure 1. Capabilities categories as reactive vs. proactive.

3

Once data is presented in this way, decision makers can conduct what-if scenarios to anticipate the full impact of the decisions that they make today. Consider, for example, the ability to assess students’ progress during the semester, which then allows advisers to intervene promptly with outreach to students who are underperforming. Using data mining, statis-tical analysis, forecasting, text analytics, and optimization and simulation, there are no limits on the insights users can gain – without relying heavily on IT.

Self-Service Analytics in Action One of the best ways higher education institutions can learn what’s possible and how to gain the most value from self-service analytics is to look at the examples set by other institutions that have already achieved success:

• NC State University is a leader in analytics use among higher learning institutions. The university had been using separate systems for student information, faculty information, finance, human resources, facilities, etc., so that users could only report on data from individual data silos. To improve its ability to manage data, NC State linked all data and put it in a usable format for self-service analysis and reporting – thus transforming decision making across the university.

• UNC-Chapel Hill has developed and implemented a new and more robust data warehouse and sophisticated data visualization solution that will help modernize the way that the university utilizes reporting.

• Des Moines Community College uses self-service reporting and analytics across the college to efficiently and effectively attract, retain and graduate students. Through self-service access, users can securely view data and reports, giving staff more time to mine data in greater detail to more proactively help students succeed.

• The University of Texas System has implemented a self-service information portal to help make informed decisions and to gauge progress toward the goals of its strategic plan. An interactive data warehouse with a user-friendly dash-board serves several purposes and has improved transpar-ency by providing open, public access to information. Now leadership can easily understand, evaluate and customize information and reports (with minimal IT reliance) to make data-driven policy decisions and assess progress.

• The University of Oregon developed a “Financial Aid Simulator” for administrators to redesign their merit aid program to recruit high achievers more effectively. They’ve uncovered deeper insights into the behavior of applicants who were accepted and offered merit aid, thus increasing the likelihood that these students will enroll. Administrators have also learned how much merit aid is needed in a finan-cial aid package to make high-achieving, in-state students more likely to enroll.

• At the University of Central Florida, administrators and managers are able to easily access and analyze informa-tion to help them make strategic, proactive decisions. The university’s information portal provides easy access to data from multiple sources, efficient self-service report genera-tion, multiyear trending and multidimensional analysis, and increased data security.

• University of Alabama uses self-service analytics to intervene quickly before at-risk students drop out. It can also identify high school students who are most likely to be interested in enrolling. This effort amounted to increased revenue through successful recruitment and retention, and in turn, helps fund capital expenses and other projects that will help attract more students.

Each of these institutions started with raw data and a vision. They’ve seen impressive success in using self-service analytics to drive change and transform their processes, operations and funding.

How SAS® Can HelpDelivering on this new paradigm, SAS Visual Analytics addresses the education-specific issues of complexity and time-liness, effectively removing the barriers to self-service analytics that have previously prevented higher education from realizing the benefits of data-driven decisions.

• Powerful advanced analytics. SAS Visual Analytics handles the once daunting issue of complexity for end users and IT alike in several ways. For end users, it provides an interface specifically designed for nonprogrammers. For example, users can easily create hierarchies (e.g., year/semester/month/day or department/faculty) and simply drag and drop variables they want to explore for trends and correlations. End users don’t need to create OLAP cubes to get the results they need, meaning no support from the IT department is required. Instead, powerful predic-tive analytics, combined with visual data exploration, are processed on the fly with in-memory computing to a wide

4

array of users – on any web interface. Users simply log on to the solution’s interactive tables and design reports exactly the way they want to see them.

• Information silos vanish. For IT, complexity often results from the problems created by disparate data sources and big data in general, since big data typically includes data from multiple sources (e.g., human resources, course data, student enrollment, facilities, etc.). SAS Visual Analytics addresses the problems arising from these disparate sources via a metadata approach. Put simply, this approach makes all the data look the same to the analytical engine, eliminating the need for complex conversions.

• Approachable analytics. Institutions can give end users access to interactive analytic visualizations, removing the previously needed – and expensive – IT overhead of data scientists and analytics experts preparing their reports. Self-service, ad hoc visual data discovery and exploration put fast insights within everyone’s reach – even those who lack analytic skills. No coding is required for even the most sophisticated analytics (e.g., decision trees, network diagrams, on-the-fly forecasting, goal seeking and scenario analysis, path analysis, and text sentiment analysis).

• Mobility. Obviously, the results of analysis need to be displayed to meet users’ needs if they are to have value, and this need complicates the challenge of enabling mobile devices, which come in a variety of shapes and sizes – not a trivial problem when visual displays are involved. Fortunately, in addition to web access, SAS Visual Analytics includes built-in native iOS and Android functionality, so IT departments don’t have to devote extra time and energy to mobility issues. As a result, advanced analytics and fast decision making are put into the hands of anyone who has secure access – at any time.

• Information sharing. How the results of analysis will be shared isn’t often perceived as a barrier to adoption, but it’s an impor-tant consideration. As IT departments have learned by long experience, users tend to cling to what’s familiar. SAS Visual Analytics is compatible with a wide variety of commonly used business tools, including Microsoft Office and SharePoint, to foster collaboration and insights on enterprisewide data consistent with users’ comfort zones.

• Flexible deployment. SAS Visual Analytics is available in multiple deployment modes: on-site, distributed and in the cloud. And SAS supports Hadoop distributed file system (Cloudera or Hortonworks distributions), as well as Teradata and Pivotal databases. With all these options, organizations can match their hardware capacity to their budget and timing needs – and easily scale it over time as needs change and grow.

• Sophisticated tools. For more advanced users, such as data scientists, SAS continues to deliver sophisticated tools in SAS Visual Analytics to ensure that your college or university will have all the analytical power you need in the future:

o Statistical analysis. Makes available quality-tested algorithms that are constantly updated to reflect the latest statistical methodologies, so you can analyze the past, describe the present and predict the future.

o Autocharting. Automatically picks the best graph.

o Powerful analytics. Provides forecasting, scenario analysis and other analytic visualizations.

o Optimization and simulation. Offers a wide range of operations research methods – including optimization, simulation and project scheduling techniques – to evaluate all alternatives.

o Text analysis. Provides insight into hot topics and includes content categorization.

o Integration with mapping technologies. Adds geospecific information to reports.

SAS self-service analytics capabilities bring powerful analytics directly to end users – with absolutely no coding required. SAS has integrated sophisticated analytics with easy-to-use features, such as autocharting, “what does it mean” pop-ups and drag-and-drop capabilities. Now anyone can gain insight and benefit from all the complex and disparate data throughout your institu-tion. For an example, Figure 2 illustrates how business users can explore data on applied, admitted and enrolled students, as well as many other combinations.

Figure 2. SAS® Visual Analytics: Student Enrollment – Applied, Admitted and Enrolled.

5

Case in Point: Western Kentucky University Gains Insight Into Its Overall PerformanceInstitutional research departments form the nucleus of data management, analysis and reporting for decision support at colleges and universities. Responsible for collecting, analyzing and reporting information about a school’s students, faculty, programs and courses, Western Kentucky University’s IR Office relies on SAS to provide university decision makers with timely access to data and information that delivers deep insight into the affairs of the institution.

It is one of the few departments on campus that incorporates different data sources, such as combining human resource data with course information. The IR team also integrated outside data sources to develop a better understanding of what it takes for a student to be successful at the university.

“The Office of Institutional Research has used SAS for 20 years to build a strong data foundation, provide analyses and meet the needs of our users,” says Tuesdi Helbig, Director of Institutional Research at Western Kentucky University. “Implementing SAS Visual Analytics is the next logical step to help us provide our users with self-service access to informa-tion. The key benefit of SAS is time savings for both the IR office and its constituents.”

Helbig says over time SAS has provided administrators and faculty members with more insight into the school’s overall performance than what they have had access to in the past, including program enrollment performance. With SAS Visual Analytics, they will be able to take this one step further by providing a highly interac-tive user experience that combines advanced data visualization, an easy-to-use interface and powerful in-memory technology. Users will be able to visually explore data, execute analytics and understand what the data means via the web, mobile devices or Microsoft Office applications.

}}“SAS is our tool – it’s our go-to solution for data, analytics and visualization.”

– Tuesdi Helbig, PhD, Director, Institutional Research

Learn More SAS solutions offer the benefits of efficient and effective use of IT resources and informed enterprisewide decisions through speed, data exploration and visualization, and deep analytics through self-service analytics. Our higher education solutions offer a powerful, flexible business analytics framework that includes capabilities for:

• University business. Analyze information on student enroll-ment, faculty, courses, finance, hiring/salary, facilities, etc., for budget alignment, accreditation or resource forecasting.

• Strategic plan development and monitoring. Slice and dice data to predict the outcomes of different scenarios in your strategic planning.

• Institutional research. Speed and improve processes and give institutional decision makers the insights they need to make proactive decisions. Use prebuilt reports to quickly and easily generate the ones most frequently needed.

• Enrollment management. Access information from across multiple sources, share information easily across institutional boundaries, and uncover patterns and trends to attract the right students, maximize retention and sustain strong rela-tionships throughout the student life cycle.

• Institutional advancement. Easily access and consolidate historical data on donors, alumni and prospects. Create predictive models that determine the likelihood of donor giving, and cost-effectively target those prospects with the highest propensity to give.

For more information on how SAS helps higher education succeed, visit: sas.com/education.

And to see a demo of a SAS Visual Analytics interactive report for higher education, or to get full access to SAS Visual Analytics, visit: sas.com/software/visual-analytics/demos/student-enrollment.html.

To contact your local SAS office, please visit: sas.com/offices

SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies. Copyright © 2014, SAS Institute Inc. All rights reserved. 107389_S130828.1214