Cloud-Based Analytics in Higher Education - Real Value or Just Pi in the Sky?
Big Data. Internet of Things. Predictive Analytics! Today you’re as likely to hear these IT buzzwords in the halls of academia as words like recruitment and retention. With more colleges and universities moving their platforms to the cloud as a cost-saving measure, there is an opportunity to glean insight from the massive amounts of data being gathered from various systems and touch points across campus—from student information systems and CRM platforms to learning management systems, student portals, social media, badges and Smartphones— to help drive student success.
While institutions understand the value of predictive analytics for driving student success in theory, many view it as yet another huge open-ended expense, especially as budgets continue to contract. This is why having a very clearvision of what your institution is trying to extract from this massive amount of data is as important as the price of the latest predictive analytics systems and real-time data crunching tools. Determining what you’re trying to gain from the data should be the first step; it also happens to be the most difficult step.
Before joining Campus Management, I worked in the energy sector in Houston, Texas. I watched the state’s economy rise, fall, and rise again with the price of oil. As we have seen recently, oil prices have dropped precipitously, displacing tens of thousands of workers in states with energy-based economies, with schools in those states now rushing to create new programs and credentials to serve them. This is where those IT buzzwords take on a profoundly human dimension. When we boil down these terms to their core value for higher education, they are essentially ways to mine the data to uncover the unique dynamics and needs of each institution and its constituents.
If I were an administrator at a college in Houston seeing massive layoffs in the energy sector, I would want to understand the backgrounds of the people affected (professional, academic, economic) and what types of programs I could create to augment their skills for related industries or even new career fields. I would want to know what the impact to operations would be if we opened a satellite campus in La Porte, where a huge number of refineries are shutting down and laying off employees. How much will I have to increase tuition to offset the drop in donations from energy companies? To what extent should I reduce enrollment in my petroleum engineering program and for how long?
To make informed program recommendations and changes in a rapidly changing climate requires not only a mash up of data extracted from the SIS, CRM, career services, alumni and advancement systems, but massive amounts of real-time economic, demographic and employment data. As more institutions recognize today, it’s equally important to leverage data from employers and state and federal workforce agencies to identify trends and align programs with the needs of the economy.
Looking Forward with Systems that Only Look Backward
ERP or SIS platforms provide an historical look at data. They tell you how many students are enrolled this semester, how many graduated last year, and how many alumni are gainfully employed. With a little help from an Excel spreadsheet and some pivot tables, they do a reasonably good job as a reporting and compliance tool using historical data. However, they can’t predict whether a prospective academic program will meet its enrollment objectives or prescribe the ideal channels for boosting enrollment, retention, outcomes, and alumni support. That’s the difference between descriptive and predictive analytics.
With higher education budgets on the decline, the goal of today’s predictive analytics initiatives is to empower executives, administrators, advisors, faculty and staff with the right information at the right time
There was a time when we depended on data scientists to understand how to extract and define data sets. Back in my energy days, we had massive IT departments with huge budgets. Not only was this expensive, it shifted responsibility and resources away from the people closest to the problem, the business units and key decision-makers who led initiatives for customer engagement and growth.
This is why predictive analytics initiatives so often failed in the past. IT departments were overwhelmed with requests from business units to cull and contextualize data from disparate systems, databases, files and web services to create accurate and meaningful reports. This also drove up the cost of IT.
With higher education budgets on the decline, the goal of today’s predictive analytics initiatives is to empower executives, administrators, advisors, faculty and staff with the right information at the right time. IT still plays an important role in facilitating strategies, but the questions now belong to decisionmakers. Instead of an IT staffer entering SQL queries like SELECT COUNT(*) FROM LEADS WHERE PROGRAM = “X”, the admissions executive is asking the same question on her laptop or tablet using a natural language query: How many leads do I have for “x” program?
While the cost of enterprise technology is a legitimate concern for any institution, especially in this current era of reduced funding, ultimately, the price of any cloudbased analytics initiative will have to be measured against the bottom-line value of the insight gained from it.
As with any critical decision, the hardest part is asking the right questions.