08-19-2026
Though it has been nearly four decades since we graduated from the Daniels (Krannert) PhD program, our appreciation for the education we received has not diminished. We had the privilege of attending one of the strongest OM programs in the country and were mentored by an outstanding faculty who trained us to look beyond the obvious. We left Purdue with very strong optimization, modelling and problem-solving skills. This training opened a number of doors for us eventually helping us serve as principal architects for Amazon’s delivery station network and UPS’s ORION system.
As good as the training was, when we stepped into the practical world we faced a jarring reality: Our PhDs hadn’t prepared us to handle real-world chaos.
During our education, the objective was well defined and the data usually pristine. In academia, if the problem is hard to formulate and solve, your advisor points out similar studies that you can leverage for ideas and support. You start assuming away troubling constraints and contradictions until the problem becomes manageable.
The real world is messy. The objectives are not clear. Very few constraints are hard and fast. There are always exceptions. Data is dirty, hard to get and has questionable sources. You must constantly figure out where data comes from, what it means, and what is broken about it. Elegant models rarely fit. As a leader, you don’t get to choose your problem and cannot "assume away" an inconvenient part of a problem to make the math easier or assumptions fit. Most importantly, you must make a decision in the time you have available and live with the consequences.
We quickly found that to make a meaningful impact, we had to get our hands dirty and learn everything from scratch. True breakthroughs came from the agonizing work of understanding the actual problem, discovering where a model adds value — and recognizing where it is useless, which data was available, and what we had to do to make the data useful. We had to understand not only what the customers thought they wanted, but also what they really needed. And more often than not, there were multiple customers at multiple levels, sometimes each with different objectives. The best lessons we learned came from our experience in intense, short-timeline startup and small business environments where our impressive academic toolkit almost felt inapplicable.
That leads us to our current perspective: business schools that best prepare the next generation of leaders championing experiential learning. Also, the current model for providing that is inadequate.
The current Fortune 500/B-school capstone project model breeds "innovation theater." Large corporations hand students low-stakes sandbox projects and at the semester's end, executives celebrate the polished presentation for a quick PR win, then quietly sweep it under the rug. The student gets an A, but never faces the ultimate operational metric: Did the solution actually work? Between us, we have seen dozens of student projects in “real world” environments, and none of them have led to lasting results. This is a disservice to both the students and the companies. The students haven’t had a “real world” experience, and the companies haven’t received value.
We propose a radical alternative: embed students (and faculty) directly into resource-constrained small businesses — because hard problems can only be solved by those affected by them.
Small businesses and startups don't have the padding for non-value-added whiteboard exercises. If a process breaks, they miss payroll. In this crucible, real problems stubbornly refuse to map to a neat course syllabus, a fixed schedule or a professor’s research. Success in this environment forces next-generation managers to build true execution grit.
We have seen evidence of the potential for this approach in the work that we have done. In one case, it was supporting a small cut & sew enterprise struggling with cash flow and operational efficiency. What appeared to be an immediate need for a materials requirements planning (MRP) system, actually required addressing fixing poor lighting, a cluttered dock and a mountain of scrap mixed in with production materials. At an aircraft manufacturer, there was no chance for successful adoption of vendor managed inventories and a Kanban system until we had physically counted every part and material in the building and identified and relocated all the just-in case materials that operators had stashed under desks, behind machines and in hallways.
In each of these cases, nothing we discussed in graduate school, presented in INFORMS meetings or built into a consulting slide deck hinted at the real and immediate problem that had to be addressed. Our real learning and contribution took place when we set aside our assumptions and models and met the businesses exactly where they are. Business schools preparing the next generation must push students off the whiteboard and onto the small business floor. That is where true operators are born.
Even in the age of AI, the immersive training remains needed and valuable.
Ranga Nuggehalli and Ed McGavin met as Purdue PhD candidates in Operations Management and worked together on their theses, discussing operations research, and the carelessly confident application of theory in real world situations. Nuggehalli spent his career at UPS as a principal scientist — co-architecting the legendary ORION system — and is an INFORMS Fellow and Certified Analytics Professional. McGavin worked as a management consultant before leading the design of Amazon’s last-mile delivery stations. In years of friendship, they’ve collected executive logistics battle scars proving that real business breakthroughs happen on the floor, not on a whiteboard.
