This blog post summarizes my top three “must know” takeaways from last week’s Higher Education Data Warehousing Forum (HEDW). My opinion is that higher education is an industry in crisis; that this audience (along with the leadership teams at most institutions) is “behind the curve” addressing these issues; and that there are concrete steps to take near-term to meet the coming challenges.
To put these takeaways in context, the event includes four days of training sessions and meetings, presentations by universities and select vendors, and a vendor showcase. The Performance Architects team was proud to attend for our tenth time this year (!) as an invited sponsor and speaker.
Takeaway #1: The U.S. higher education market is late to the table preparing for massive shifts coming in the next few years and needs to get ready to address these changes; the HEDW community is uniquely positioned to help institutions use data to weather this storm.
The U.S. higher education market is facing a major supply and demand issue, as well as a disintermediation of its business model (like Uber and taxis, or the VCR and streaming video) and the HEDW crowd was, surprisingly, almost silent about these changes. The HEDW community is comprised of the people at institutions best-positioned to use data to address these challenges, and I urge the HEDW and larger higher education community to start focusing on these issues!
In recent decades, the higher education market was one of the U.S.’s unstoppable growth sectors as a result of an explosion in the college-age student population. The National Center for Education Statistics notes that enrollment grew an incredible 54% from 1990 to 2011, and the number of institutions grew 30% between 1980 and 2015 to meet this demand.
The issue is that enrollment is expected to shrink at least 15% after 2025, because the U.S. is running out of teenagers! The share of recent high school graduates going on to college has barely budged, with fewer recent high school grads overall due to declining birth rates.
In addition, the perceived value of a degree has evolved drastically.
It costs more to get a degree now than it ever has, and these increases don’t seem to be leveling off anytime soon. The U.S. Bureau of Labor Statistics notes that the increase in tuition at an average U.S. institution went up ~732% from 1983 to 2013.
According to the National Association of College and University Business Officers (NACUBO), funding for institutions shifted during the Great Recession to a primarily enrollment-driven model because states, on average, cut their funding ~30%, and institutions shifted these costs to families and students. Although the average student at a “top 20” (Ivy or Ivy-equivalent) institution is only paying 15% of the total cost of their education, the average student at a “regular” private college (all private colleges except the top 20 Ivy or Ivy-equivalent) is paying 75% of the total cost of their education, meaning that most private colleges are struggling to attract students who can carry these costs.
At the same time, the value of a degree just isn’t the same. 2011 was the year that student loan indebtedness surpassed credit card indebtedness in the U.S. for the first time, and ~40% of college grads in 2018 took (or were forced to take) a job that didn’t require a college degree and were – to some extent – “underemployed” (important note: there’s lots of debate about these statistics).
As a result, we’re moving to a “cost/benefit” driven perspective: Studies demonstrate a significant shift in attitudes from “Everyone needs a degree” to “What is the cost/benefit for a degree for my line of work?”
Where does this leave us? Institutions must get a handle on their financial picture or else they’re going to go out of business…and this requires accurate, complete data and information! The price for an average U.S. institution to deliver a degree grew 233% from 1983 to 2013. Of ~800 U.S. schools surveyed by NACUBO in 2017, the top 50 wealthiest had a median endowment of $3.5B, but for the entire list, the median was just $113M. The same study also outlined that >30% of institutions are spending more than they can afford to spend and >70% of this cost is salary and benefits for an average institution.
The scary thing is, they’ve already started to go out of business! I live in Boston, and the February 2019 issue of Boston Magazine contains a cover story called “The End.” The article discusses how small, private colleges are going bankrupt and are causing a crisis for the city and their constituents. I think this is only the tip of the iceberg, with more to come.
Takeaway #2: Heterogeneous, hybrid analytic environments are going to be the norm for the foreseeable future.
HEDW is a vendor-agnostic conference, meaning that attendees and presenters have the freedom to discuss whatever they’d like, with the result that the most frequently used or popular solutions tend to “bubble to the surface” in speaking topics and conversations.
This year, it was evident that no one vendor in any analytic stack category (including data integration and governance, modeling, analytics/BI, etc.) really stood out. The theme of conversations was how to best address institutions’ needs in a hybrid on-premises/cloud world, where multiple “best of breed” solutions will be used in each area. As one speaker put it, “We don’t know who’s going to win the race, so we think of it like a toolbox. You don’t just want one tool in your toolbox, you want lots of different tools optimized for certain purposes.”
Takeaway #3: Data integration, data governance, scalability, and security are critical to succeeding in this hybrid, heterogeneous world.
The issue and opportunity with this heterogeneous world is that organizations can no longer govern data definitions and integration the way they could when most of their data resided behind a firewall, and HEDW Forum attendees commented that this was one of the largest analytics issues facing their institutions today.
In Performance Architects’ presentation, “Data Integration in the Cloud Era: Case Studies, Tips, and Tricks on How to Succeed” (currently available for free on the Performance Architects Learning Center here; sign up is required but it’s free to join!), we discussed these issues and proposed some actions to take short term to address these items, including how to develop a data integration and governance roadmap and organization as well as what data and processes we recommend leaving on-premises or moving to the cloud nearer-term.
Scalability and security were the other major themes that require attention. Scalability is an issue because many of the “self-service” visualization solutions that have become popular in the past few years are simply not designed to address the data volume that’s required for a large enterprise analytics effort. Security is also a concern because, again, many of these solutions were not built with large work group security needs in mind (e.g., single sign-on or SSO or row-level or user-based security capabilities).
How HEDW Members Can Act on Event Takeaways
The HEDW community should be alarmed by the huge changes taking place in their market, and they need to act now to prevent their organizations from going the way of the taxi or the VCR. Since the HEDW community includes analytics “thought leaders” across U.S. institutions, this group has the power to help their institutions’ leadership teams turn data into information to address their business concerns.
First, this group needs to help institutions stabilize their analytics environments. This means coming up with data governance, integration, scalability and security capabilities that will persist in this heterogeneous, hybrid world that is now the norm.
Second, this community needs to help frame the conversation by turning data into information that helps address the trends that threaten to turn the U.S. higher education market into a “has been.” This means helping institutions get a handle on how to control expenses; how to shift the composition of the student population; and how to possibly alter revenue sources to address potential shortfalls.
Finally, newer analytics capabilities that are just starting to become “mainstream” such as machine learning (ML) and artificial, adaptive or autonomous intelligence (AI) may offer interesting and novel ways to investigate data to address some of these concerns…but that’s the topic for another blog post (for an introduction to some of these concepts, go to Performance Architects’ recent blog post entitled, “Buzzword Breakdown – Blockchain, Big Data, IoT, Machine Learning, AI Webinar Preview” or to the webinar slides on our Learning Center).
I also want to thank the University of Michigan team for a fun and well-run conference. We had a blast touring “The Big House” (the Michigan stadium). These are some of our favorite photos of the event. First, the now-annual ice sculpture we believe started at Penn a few years back:
A view of the stadium from the clubhouse level (did you know it seats 107,000+ people?!):
Storming the field with conference attendees:
An inspiring message from the tunnel leading onto the field:
If you’re interested in assistance with any of the takeaway topics covered in this blog post, please contact us at email@example.com and we would be happy to set up a time to discuss.