What Nine Years of Tracking ISM’s Top 30 Under 30 Reveals About Why Procurement Implementations Keep Failing
Procurement Insights — Jon Hansen | March 2026
In 2017, ISM and Thomas recognized 30 rising stars in supply chain and procurement — the profession’s best and brightest under 30. These weren’t theoretical picks. They were practitioners. Category managers at Johnson & Johnson. Supply chain analysts at Google. Procurement supervisors at ExxonMobil. Commodity managers at Dell, Boeing, and US Steel.
Nine years later, I tracked all 30.
The career data is impressive. But what it reveals — when you hold it against the industry’s persistent implementation failure rate — is devastating. Not for the individuals. For the system they’ve been working inside.
Part One: What Five AI Models Agreed On (and Where They Didn’t)
Before presenting the findings, I want to show you how they were validated — because methodology matters, and the process itself tells a story.
Using the RAM 2025™ Multimodel/Multilevel Validation Framework, I ran the same research question through five independent AI models: Where are the 2017 ISM/Thomas 30 Under 30 now, and are they still in procurement?
The models didn’t coordinate. They didn’t see each other’s work. They used whatever public sources they could find — LinkedIn, ZoomInfo, company profiles, industry press, ISM publications — and they returned five independent assessments.
Here’s what happened.
The Three Tiers
Tier 1 — Evidence-Anchored, Limits Acknowledged
Two models confirmed 8–13 individuals with specific career paths, employer names, and source attribution. When they couldn’t verify someone, they said so. When a data point was thin, they flagged it. These models treated “I don’t know” as a finding, not a failure.
Tier 2 — High Confidence, Thin Sourcing
Two models claimed 26–27 of 30 were still in the profession, with retention rates of 87–90%. They provided names, titles, and employers — some of which no other model corroborated. One listed an individual as “Director, Materials Management” at Hershey; the evidence-anchored model placed the same person as “Senior Manager, Supply Insights & Analytics.” One claimed a cohort member was at Zoox; no other model found this. Another placed someone at Carpenter Technology with no corroboration.
The pattern: highest confidence, lowest corroboration. In the RAM 2025™ framework, that is a hallucination signature.
Tier 3 — Methodological Refusal
One model declined to answer entirely. It said the task couldn’t be done reliably with automated tools and offered to build a spreadsheet framework for manual research instead.
This refusal is not a weakness. It’s a data point. When a model tells you “this question can’t be answered with the tools available,” and two other models answer it confidently anyway, that divergence is the finding.
What Converged
Across all five models, 12–14 individuals were independently confirmed by two or more models with matching employer and role data:
Charlotte de Brabandt — now Dr. de Brabandt, independent keynote speaker and digital futurist. Mitchell Agee — Senior Manager, F-35 Customer Programs (Japan), Lockheed Martin. Manuel Allendesalazar — Associate Director, Demand Planning, Teva Pharmaceuticals (same employer since 2017). Ajay Kumar Arjunan — PMO Lead, Global Supply Chain, Dell Technologies (same employer since 2017). Neta Berger — Operations Program Manager, Google (same employer since 2017). Steven Clowney — Senior Manager, Indirect Sourcing (IT), Genuine Parts Company. Rhiana Gallen — Director, ProcureAbility. Jordan Haller — Senior Vice President, Procurement and Supply Chain, QTS Data Centers. Caitlin Michaelis — Director, Partner Management Governance, Unum. Tanner Ryan — Senior Category Activation Manager, Danone. Keith Unton — Senior Manager, Supply Insights & Analytics, The Hershey Company. Conor Quarry — procurement analytics, IBM. Chelsey Graham — Manager, Solution Consulting, Ivalua (moved from practitioner at US Steel to procurement-tech vendor side). Daniel Sanchez — category management, ISM Boston chapter leadership.
Of those confirmed: 100% are still in procurement, supply chain, or a directly adjacent function.
Not one left the profession.
What the RAM 2025™ Process Exposed
The defensible claim is not “87–90% retention.” That number comes from models filling gaps with confident-sounding but unverifiable specifics.
The defensible claim is: Among the 12–14 independently confirmed across multiple models, retention is 100%. The true number still in the profession is very likely higher — possibly 25 or more — but the gap between “likely” and “confirmed” is exactly the kind of distinction that separates evidence-based analysis from analyst theater.
This is what multimodel validation does. It doesn’t just find answers. It finds the boundary of what’s knowable — and it shows you where confidence outruns evidence.
Part Two: The Convergence That Matters
Now here’s where the data gets uncomfortable.
These 30 people were recognized in 2017 because they were exceptional. ISM and Thomas didn’t select them randomly. They were chosen for early career impact, leadership potential, and professional commitment. And the nine-year tracking data confirms the selection was sound — these individuals stayed, advanced, and in many cases reached Director and VP-level positions at major organizations.
Consider what this cohort has collectively experienced since 2017:
Technology eras traversed: Cloud procurement platforms (mature by 2017), AI-augmented sourcing (2019–2022), generative AI disruption (2023–2024), and agentic AI emergence (2025–2026). Four distinct technology waves in nine years.
Disruptions survived: COVID-19 supply chain collapse, the Great Resignation, semiconductor shortages, geopolitical supply fragmentation, inflation-driven cost pressure, and the AI talent war. Any one of these would test a career. This cohort navigated all of them — and stayed.
Career progression achieved: Among the confirmed group, four reached Director level or above. Jordan Haller made SVP at QTS Data Centers. Rhiana Gallen — the youngest winner at 23 — is now a Director at ProcureAbility. Keith Unton completed 11+ distinct roles at Hershey spanning every major supply chain function. Charlotte de Brabandt earned a doctorate and became one of procurement’s most recognized public voices.
The talent pipeline worked. ISM’s selection criteria identified people who would lead.
And yet.
The procurement technology implementation failure rate in 2026 is functionally identical to what it was in 2017. And in 2012. And in 2007. Seventy to eighty percent of implementations fail to deliver their projected outcomes. The number has not moved.
Part Three: The Question Nobody Is Asking
If the talent stayed — and it did — and the technology improved — and it has — then why hasn’t the failure rate changed?
This is the question the industry keeps not asking. Vendors don’t ask it because the answer implicates their go-to-market model. Analysts don’t ask it because their revenue depends on technology selection guidance. And procurement leaders don’t ask it because the answer suggests the problem isn’t the tool, the team, or the budget.
The problem is when you start.
These 30 individuals — talented, committed, ISM-recognized — have spent nine years working inside organizations that buy technology before they assess whether they’re ready to use it. They’ve implemented platforms that were selected before anyone measured the behavioral, cultural, and process gaps that determine whether a deployment succeeds or fails.
They didn’t fail. The system they operate in was never designed to succeed at the point it matters most: before the technology decision is made.
This is Phase 0.
The Three Archetypes and What They Tell Us
The career tracking revealed three distinct patterns among the confirmed cohort:
The Loyalists — Unton, Allendesalazar, Arjunan, Berger. Stayed with their 2017 employer. Advanced internally. Unton’s 11 internal moves at Hershey represent the most complete supply chain generalist arc in the dataset — distribution, production, demand planning, external manufacturing, strategic sourcing, SAP implementation, analytics. He has seen every function. He knows where implementations break.
The Climbers — Haller, Gallen, Michaelis. Changed companies strategically, each move upward. They leveraged industry transitions as accelerators. Haller went from ExxonMobil to Louisiana-Pacific to QTS Data Centers. Every move was a step up — and every move put him inside a new organization’s implementation reality.
The Broadeners — de Brabandt, Agee, Clowney, Ryan. Moved across industries, applying procurement skills in new contexts. De Brabandt went from J&J to ZF Group to Amazon to independent practice. Clowney moved from food service to agriculture to automotive parts. They’ve seen procurement from the inside of multiple organizational cultures.
Every archetype tells the same story from a different angle. The Loyalists saw it from depth: the same organization cycling through technology waves without fundamentally changing how it prepares. The Climbers saw it from altitude: rising high enough to see the pattern repeat across companies. The Broadeners saw it from breadth: the same readiness gaps showing up in completely different industries.
The failure rate isn’t a technology problem. It’s not a talent problem. It’s a sequencing problem. And this cohort — collectively — has nine years of lived evidence that proves it.
What This Means
The ISM/Thomas 30 Under 30 Class of 2017 did everything right. They stayed in the profession. They earned certifications, MBAs, and doctorates. They took on harder roles. They led through disruption. They are, by any reasonable measure, exactly the kind of professionals the industry said it needed.
And the failure rate didn’t move.
The default industry narrative is always talent-facing: we need better people, better training, better certifications, better leadership pipelines. ISM built the 30 Under 30 program on that premise. And the nine-year data proves the premise worked — the talent showed up, stayed, and advanced.
But the system these people entered was already broken before they arrived. They were handed implementations that skipped readiness assessment. They were deployed into organizations that selected technology before measuring capability gaps. They were asked to make platforms succeed in environments that were never diagnosed.
Keith Unton — 11 roles at Hershey, every supply chain function, SAP S/4 implementation lead. If anyone could muscle an implementation to success through sheer breadth of knowledge, it would be him. But one person’s capability cannot override an organizational readiness gap that was never measured. The system absorbs the talent without changing the outcome.
Jordan Haller — SVP at 37, three industries, MBA, sustainability certification. He has the authority, the scope, and the cross-industry perspective. And he is still operating inside the same vendor-first, readiness-last procurement cycle that existed when he was a supervisor at ExxonMobil in 2017.
The question that emerges from this data is not whether the talent is failing the industry. The talent did its part. The question is whether the industry is failing its best talent.
The answer, based on nine years of evidence, is yes.
The industry invested in these people and then put them inside a process designed to waste that investment. It identified rising stars and handed them structurally compromised implementations. It celebrated their potential and then deployed them into environments where the outcome was already determined — not by their skills, but by the readiness gaps no one measured before the purchase order was signed.
The Hansen Fit Score™ exists because of this pattern. RAM 2025™ exists because of this pattern. And Phase 0 exists because of this pattern — not just as an organizational diagnostic, but as a talent protection strategy. Without it, the industry will keep burning through its best people on implementations that were structurally compromised before day one.
For the 2017 cohort — now 32 to 39, entering the peak leadership years — the next five years will determine whether they inherit the same broken system or help build the one that replaces it.
The talent was never the problem. Wasting it was.
This research was validated using the RAM 2025™ Multimodel/Multilevel Validation Framework across five independent AI models. The full methodology, career tracker, and model-by-model assessment are available through Procurement Insights.
Previous coverage: “The ISM ThomasNet Top 30 Under 30 Class of 2017: Where Are They Now?” — Procurement Insights, January 2024
-30-
Procurement Insights
The Best Talent. The Best Technology. The Same 70% Failure Rate.
Posted on March 3, 2026
0
What Nine Years of Tracking ISM’s Top 30 Under 30 Reveals About Why Procurement Implementations Keep Failing
Procurement Insights — Jon Hansen | March 2026
In 2017, ISM and Thomas recognized 30 rising stars in supply chain and procurement — the profession’s best and brightest under 30. These weren’t theoretical picks. They were practitioners. Category managers at Johnson & Johnson. Supply chain analysts at Google. Procurement supervisors at ExxonMobil. Commodity managers at Dell, Boeing, and US Steel.
Nine years later, I tracked all 30.
The career data is impressive. But what it reveals — when you hold it against the industry’s persistent implementation failure rate — is devastating. Not for the individuals. For the system they’ve been working inside.
Part One: What Five AI Models Agreed On (and Where They Didn’t)
Before presenting the findings, I want to show you how they were validated — because methodology matters, and the process itself tells a story.
Using the RAM 2025™ Multimodel/Multilevel Validation Framework, I ran the same research question through five independent AI models: Where are the 2017 ISM/Thomas 30 Under 30 now, and are they still in procurement?
The models didn’t coordinate. They didn’t see each other’s work. They used whatever public sources they could find — LinkedIn, ZoomInfo, company profiles, industry press, ISM publications — and they returned five independent assessments.
Here’s what happened.
The Three Tiers
Tier 1 — Evidence-Anchored, Limits Acknowledged
Two models confirmed 8–13 individuals with specific career paths, employer names, and source attribution. When they couldn’t verify someone, they said so. When a data point was thin, they flagged it. These models treated “I don’t know” as a finding, not a failure.
Tier 2 — High Confidence, Thin Sourcing
Two models claimed 26–27 of 30 were still in the profession, with retention rates of 87–90%. They provided names, titles, and employers — some of which no other model corroborated. One listed an individual as “Director, Materials Management” at Hershey; the evidence-anchored model placed the same person as “Senior Manager, Supply Insights & Analytics.” One claimed a cohort member was at Zoox; no other model found this. Another placed someone at Carpenter Technology with no corroboration.
The pattern: highest confidence, lowest corroboration. In the RAM 2025™ framework, that is a hallucination signature.
Tier 3 — Methodological Refusal
One model declined to answer entirely. It said the task couldn’t be done reliably with automated tools and offered to build a spreadsheet framework for manual research instead.
This refusal is not a weakness. It’s a data point. When a model tells you “this question can’t be answered with the tools available,” and two other models answer it confidently anyway, that divergence is the finding.
What Converged
Across all five models, 12–14 individuals were independently confirmed by two or more models with matching employer and role data:
Charlotte de Brabandt — now Dr. de Brabandt, independent keynote speaker and digital futurist. Mitchell Agee — Senior Manager, F-35 Customer Programs (Japan), Lockheed Martin. Manuel Allendesalazar — Associate Director, Demand Planning, Teva Pharmaceuticals (same employer since 2017). Ajay Kumar Arjunan — PMO Lead, Global Supply Chain, Dell Technologies (same employer since 2017). Neta Berger — Operations Program Manager, Google (same employer since 2017). Steven Clowney — Senior Manager, Indirect Sourcing (IT), Genuine Parts Company. Rhiana Gallen — Director, ProcureAbility. Jordan Haller — Senior Vice President, Procurement and Supply Chain, QTS Data Centers. Caitlin Michaelis — Director, Partner Management Governance, Unum. Tanner Ryan — Senior Category Activation Manager, Danone. Keith Unton — Senior Manager, Supply Insights & Analytics, The Hershey Company. Conor Quarry — procurement analytics, IBM. Chelsey Graham — Manager, Solution Consulting, Ivalua (moved from practitioner at US Steel to procurement-tech vendor side). Daniel Sanchez — category management, ISM Boston chapter leadership.
Of those confirmed: 100% are still in procurement, supply chain, or a directly adjacent function.
Not one left the profession.
What the RAM 2025™ Process Exposed
The defensible claim is not “87–90% retention.” That number comes from models filling gaps with confident-sounding but unverifiable specifics.
The defensible claim is: Among the 12–14 independently confirmed across multiple models, retention is 100%. The true number still in the profession is very likely higher — possibly 25 or more — but the gap between “likely” and “confirmed” is exactly the kind of distinction that separates evidence-based analysis from analyst theater.
This is what multimodel validation does. It doesn’t just find answers. It finds the boundary of what’s knowable — and it shows you where confidence outruns evidence.
Part Two: The Convergence That Matters
Now here’s where the data gets uncomfortable.
These 30 people were recognized in 2017 because they were exceptional. ISM and Thomas didn’t select them randomly. They were chosen for early career impact, leadership potential, and professional commitment. And the nine-year tracking data confirms the selection was sound — these individuals stayed, advanced, and in many cases reached Director and VP-level positions at major organizations.
Consider what this cohort has collectively experienced since 2017:
Technology eras traversed: Cloud procurement platforms (mature by 2017), AI-augmented sourcing (2019–2022), generative AI disruption (2023–2024), and agentic AI emergence (2025–2026). Four distinct technology waves in nine years.
Disruptions survived: COVID-19 supply chain collapse, the Great Resignation, semiconductor shortages, geopolitical supply fragmentation, inflation-driven cost pressure, and the AI talent war. Any one of these would test a career. This cohort navigated all of them — and stayed.
Career progression achieved: Among the confirmed group, four reached Director level or above. Jordan Haller made SVP at QTS Data Centers. Rhiana Gallen — the youngest winner at 23 — is now a Director at ProcureAbility. Keith Unton completed 11+ distinct roles at Hershey spanning every major supply chain function. Charlotte de Brabandt earned a doctorate and became one of procurement’s most recognized public voices.
The talent pipeline worked. ISM’s selection criteria identified people who would lead.
And yet.
The procurement technology implementation failure rate in 2026 is functionally identical to what it was in 2017. And in 2012. And in 2007. Seventy to eighty percent of implementations fail to deliver their projected outcomes. The number has not moved.
Part Three: The Question Nobody Is Asking
If the talent stayed — and it did — and the technology improved — and it has — then why hasn’t the failure rate changed?
This is the question the industry keeps not asking. Vendors don’t ask it because the answer implicates their go-to-market model. Analysts don’t ask it because their revenue depends on technology selection guidance. And procurement leaders don’t ask it because the answer suggests the problem isn’t the tool, the team, or the budget.
The problem is when you start.
These 30 individuals — talented, committed, ISM-recognized — have spent nine years working inside organizations that buy technology before they assess whether they’re ready to use it. They’ve implemented platforms that were selected before anyone measured the behavioral, cultural, and process gaps that determine whether a deployment succeeds or fails.
They didn’t fail. The system they operate in was never designed to succeed at the point it matters most: before the technology decision is made.
This is Phase 0.
The Three Archetypes and What They Tell Us
The career tracking revealed three distinct patterns among the confirmed cohort:
The Loyalists — Unton, Allendesalazar, Arjunan, Berger. Stayed with their 2017 employer. Advanced internally. Unton’s 11 internal moves at Hershey represent the most complete supply chain generalist arc in the dataset — distribution, production, demand planning, external manufacturing, strategic sourcing, SAP implementation, analytics. He has seen every function. He knows where implementations break.
The Climbers — Haller, Gallen, Michaelis. Changed companies strategically, each move upward. They leveraged industry transitions as accelerators. Haller went from ExxonMobil to Louisiana-Pacific to QTS Data Centers. Every move was a step up — and every move put him inside a new organization’s implementation reality.
The Broadeners — de Brabandt, Agee, Clowney, Ryan. Moved across industries, applying procurement skills in new contexts. De Brabandt went from J&J to ZF Group to Amazon to independent practice. Clowney moved from food service to agriculture to automotive parts. They’ve seen procurement from the inside of multiple organizational cultures.
Every archetype tells the same story from a different angle. The Loyalists saw it from depth: the same organization cycling through technology waves without fundamentally changing how it prepares. The Climbers saw it from altitude: rising high enough to see the pattern repeat across companies. The Broadeners saw it from breadth: the same readiness gaps showing up in completely different industries.
The failure rate isn’t a technology problem. It’s not a talent problem. It’s a sequencing problem. And this cohort — collectively — has nine years of lived evidence that proves it.
What This Means
The ISM/Thomas 30 Under 30 Class of 2017 did everything right. They stayed in the profession. They earned certifications, MBAs, and doctorates. They took on harder roles. They led through disruption. They are, by any reasonable measure, exactly the kind of professionals the industry said it needed.
And the failure rate didn’t move.
The default industry narrative is always talent-facing: we need better people, better training, better certifications, better leadership pipelines. ISM built the 30 Under 30 program on that premise. And the nine-year data proves the premise worked — the talent showed up, stayed, and advanced.
But the system these people entered was already broken before they arrived. They were handed implementations that skipped readiness assessment. They were deployed into organizations that selected technology before measuring capability gaps. They were asked to make platforms succeed in environments that were never diagnosed.
Keith Unton — 11 roles at Hershey, every supply chain function, SAP S/4 implementation lead. If anyone could muscle an implementation to success through sheer breadth of knowledge, it would be him. But one person’s capability cannot override an organizational readiness gap that was never measured. The system absorbs the talent without changing the outcome.
Jordan Haller — SVP at 37, three industries, MBA, sustainability certification. He has the authority, the scope, and the cross-industry perspective. And he is still operating inside the same vendor-first, readiness-last procurement cycle that existed when he was a supervisor at ExxonMobil in 2017.
The question that emerges from this data is not whether the talent is failing the industry. The talent did its part. The question is whether the industry is failing its best talent.
The answer, based on nine years of evidence, is yes.
The industry invested in these people and then put them inside a process designed to waste that investment. It identified rising stars and handed them structurally compromised implementations. It celebrated their potential and then deployed them into environments where the outcome was already determined — not by their skills, but by the readiness gaps no one measured before the purchase order was signed.
The Hansen Fit Score™ exists because of this pattern. RAM 2025™ exists because of this pattern. And Phase 0 exists because of this pattern — not just as an organizational diagnostic, but as a talent protection strategy. Without it, the industry will keep burning through its best people on implementations that were structurally compromised before day one.
For the 2017 cohort — now 32 to 39, entering the peak leadership years — the next five years will determine whether they inherit the same broken system or help build the one that replaces it.
The talent was never the problem. Wasting it was.
This research was validated using the RAM 2025™ Multimodel/Multilevel Validation Framework across five independent AI models. The full methodology, career tracker, and model-by-model assessment are available through Procurement Insights.
Previous coverage: “The ISM ThomasNet Top 30 Under 30 Class of 2017: Where Are They Now?” — Procurement Insights, January 2024
-30-
Share this:
Related