Machine Learning Times
Machine Learning Times
EXCLUSIVE HIGHLIGHTS
SHARE THIS:

3 weeks ago
What Every Analytics Manger Should Know About Working With Universities. But Doesn’t.

 

 

This is part 1 of a 5-part series on university/corporate partnerships in analytics and data science

How many .edu addresses do you have in your email contacts?

If you are like most Analytics Managers, the answer is “not many”.

Academia and the private sector have historically rarely intersected.  But increasingly there is a permeable membrane between the two, with arguably more collaboration occurring than at any other time in history. Consider the following:

  • Before 2005 there were no formal university programs in the country with the title “Data Science”, or “Business Analytics”. Largely in response to the demands of the private sector, by 2019 there were over 250 programs just at the masters level and almost 100 at the Ph.D. level.
  • Almost half of all individuals earning Ph.D.’s in computational disciplines enter the private sector after graduation.
  • Multi-organization innovation and research labs, which include university faculty, graduate students, private sector R&D teams and public sector policy contributors are increasingly common, where many of these labs publish more “academic” papers than individual academic institutions. In 2017, a quarter of all “academic” publications in data science journals included an author from the private sector with no academic affiliation.
  • In 2019, Google announced that they would be conferring Ph.D. degrees.

At our own universities, we regularly see healthcare systems, financial service providers, manufacturing firms competing to hire the same student.  At first glance, that makes no sense – after all, healthcare, finance and manufacturing are completely different domains.  Just ten years ago, it was unheard of to see a big bank and a healthcare provider chasing the same talent.  Today, its commonplace.  Why?  Because while domain expertise is important, it is frequently subordinated; the bigger challenge is finding “data natives” who have the facility to work in a multi-faceted and complex data environment instead of finding a student with banking experience.

Importantly, universities are also experiencing unforeseen challenges in analytics.  Unlike more traditional disciplines like mathematics or English, the academic location of analytical disciplines within a university is not consistently defined – should it be housed in Computer Science?  Statistics?  Business?  Or in an interdisciplinary center or institute?  Analytical curriculums still have little consistency.  However, what is consistently recognized across the academic ecosystem is that analytics – like Accounting, Medicine and Engineering – is really a discipline that is best learned through application.  While most universities with analytical programs have some requirement for experiential learning (e.g., applied projects, capstone courses, internships), aligning these initiatives with the needs of a company can present challenges. Additionally, with the relatively new phenomena of individuals with PhDs pursuing careers in the private sector – while still engaging in meaningful and relevant research, AND making FAR more money than they would in academia – universities are experiencing their own “talent gap”. 

As university professors and academic administrators with both private and public sector experiences in healthcare, policy, consulting and finance, who have run university analytics programs, we have had a front row seat in the evolution of the university-private sector relationship. In our collective 40 years of academic experience (Bob is older), we have identified three broad objectives that companies – particularly those on the lower end of the analytics maturity continuum – have for partnering with universities in the context of analytics and data science:

  1. Talent Pipelines. Companies establish strategic partnerships with universities to develop pipelines of analytical talent, which goes beyond just posting on job boards or attending career fairs.      
  2. Alternative Insights and Thought Leadership. As organizational leaders consider issues related to integrating analytics more broadly to support their objectives, data science research labs and capstone projects enable innovation outside the constraints of an organization’s operations.  Research products from these labs, which integrate faculty, doctoral students and R&D teams, include collaborative publications, contributing to the company’s position as a forward-thinking industry thought leader.  These engagements frequently generate unexpected and valuable insights.  In addition, universities provide a natural infrastructure for ongoing training and executive education in data science.   
  3. Community Engagement. Hiring from local universities – and partnering with them to develop or pivot curriculum if needed to meet market demand – creates benefits to the local and regional business community that extends well beyond just new hires.  To local companies, Jennifer (based in Atlanta) frequently says “If you are going outside of Atlanta to hire analytical talent…shame on you.  Eat Local, Buy Local, Hire Local”.  

For a collaboration to be successful, both parties need to see benefits; organizational leaders need to also consider what the university will need to deem the collaboration successful.  From our experiences as faculty and as academic administrators, universities have four broad objectives for partnering with companies in the context of analytics and data science:

  1. Experiential Learning. Data science is an applied discipline.  While students can access publicly available data (e.g. Kaggle), there is no substitution for having an executive come into a classroom and say, “Here is some of my data…here are my pain points…here is what I am trying to achieve”.   Those types of experiences not only make for a more meaningful analytical learning environment, they also ultimately contribute to a more prepared graduate who can add value to their new employer on day one.  In addition, for analytics programs that are “domain agnostic”, experiential learning is an important opportunity for students to gain important domain expertise.    
  2. Research. For faculty and doctoral students, peer-reviewed publications are the “coin of the realm”; the ability to produce meaningful, innovative, relevant, independent research is a set of skills which increasingly has value in both academia as well as in the private sector.  While there are issues related to intellectual property rights and non-disclosures that need to be resolved, these issues are almost never insurmountable when the objectives of both parties are well-articulated and understood.
  3. External Funding.  
  4. Community Engagement. Universities – particularly public universities – have a fiduciary responsibility to not only produce talent that can meet the needs of the state economy, but increasingly to “use data science for social good” in the local community.      

Over the next five months, we will continue this series for analytics managers considering partnerships with universities, with specific examples and cautionary tales.

About the Authors

Dr. Jennifer Lewis Priestley is the Executive Director of the Analytics and Data Science Institute at Kennesaw State University.  Dr. Robert McGrath is the Chair of Health Management and Policy at the University of New Hampshire.

One thought on “What Every Analytics Manger Should Know About Working With Universities. But Doesn’t.

  1. Pingback: Data Science newsletter – January 29, 2020 | Sports.BradStenger.com

Leave a Reply

Pin It on Pinterest

Share This