For thirty years, the post-mortem has landed in the same place — and in the wrong office.
The ERP implementation underperforms. The eProcurement rollout stalls. The P2P platform goes live and the adoption numbers disappoint. The AI layer produces confident recommendations that do not convert into outcomes. And the narrative — in boardrooms, in analyst reports, in industry conference keynotes — assigns the failure to procurement.
Wrong function. Wrong diagnosis. Same result, every time.
We have been treating a capital allocation and governance failure as a procurement failure.
Who actually owns the failure
The CIO approved the architecture. The CFO approved the capital. The technology selection was made inside an evaluation framework designed by vendor-aligned analysts whose research was funded by the platforms being assessed. The implementation was handed to a procurement function that had no authority over the governance architecture it was being asked to operate within.
When the implementation failed — and across seven consecutive technology eras, it failed to meet expected outcomes between 60 and 85 percent of the time — consistently, across independent analyses and 28 years of archive documentation — procurement absorbed the accountability.
The function with the least authority inherited the consequences of decisions made by the functions with the most.
How do you sign off on a multimillion-dollar stack and not own it when it goes down the crapper?
That is what every practitioner sitting through a post-mortem has wanted to say out loud for thirty years.
That is not a procurement failure. That is a capital allocation failure. It is a governance architecture failure. It is an IT investment failure. And the CIOs and CFOs who have been nodding along at procurement conferences while the CPO presented the post-mortem have been watching the wrong function take responsibility for their decisions for three decades.
What the market is telling you right now
The procurement industry is contracting. Membership bodies are tightening focus. Community platforms are consolidating. Advisors who built their careers around procurement are quietly moving toward AI ventures and capital-backed platforms — toward the functions where transformation budget and decision authority actually sit.
This is not coincidence. This is the market correcting for a structural reality that has been true since 1995: procurement can diagnose the problem. Procurement cannot fund the solution. The mandate and the budget to fix structural failure sit with the CIO and the CFO — and they always have.
Deloitte’s CPO surveys have been documenting the downstream effects of this for years. More than 70% of CPOs report difficulty attracting and retaining talent capable of delivering strategic outcomes. The confidence gap between high-performing and low-performing procurement organizations is not primarily a training problem — it is an investment problem. Organizations that treat procurement as a strategic function develop their people accordingly. Organizations that treat procurement as a cost center hand their CPO a strategic mandate and a tactical budget.
The 2007 version of this problem was visible in the archive. The procurement profession’s strategic ambitions were outpacing executive willingness to invest in, empower, and retain the talent needed to produce strategic outcomes. Deloitte did not overturn that finding in 2023. They measured it.
The profession has been asking for a seat at the table for thirty years. The CIO and CFO have been handing them the chair and keeping the keys.
Why the current AI cycle is different — and more dangerous
Every previous technology era produced visible failure signals. A poorly configured ERP produced error messages. A poorly deployed P2P platform produced process bottlenecks practitioners could identify and report.
A poorly deployed AI procurement platform produces confident, fluent, coherent wrong answers — and the organization consuming those outputs has no native mechanism to distinguish them from correct ones. The danger is not that AI is mystical. It is that its mistakes are harder to see, arrive faster, and are easier for executives to rationalize. The MoE-based platforms that CIOs are currently approving and CFOs are currently funding are optimized to produce the most coherent answer from within their own data environment. They are not designed to tell you what they do not know. They are not designed to produce a Stop signal. And the analyst firms evaluating them — funded by the vendors being assessed — are not designed to produce one either.
The CIO who approved the architecture is accountable for what that architecture cannot do. The CFO who approved the capital is accountable for the ROI that architecture cannot deliver. Procurement will be handed the implementation. Procurement will absorb the post-mortem.
Unless something changes before the deployment begins.
The architectural question CIOs are not asking
The platforms are more capable than they have ever been. That is true and the CIOs who approved them are not wrong about what the technology can do in a controlled evaluation environment.
The question that is not being asked — by the CIO, by the CFO, by any vendor whose commercial model depends on the answer being proceed — is whether the organization’s governance architecture is designed to absorb what the platform is about to recommend.
Not whether the platform is capable. Not whether the data is clean. Not whether the integration is complete. Whether the human governance layer — the decision authority, the approval thresholds, the process integrity, the named individual who can act on a recommendation within the 30-day window that determines whether the response is a decision or a consequence — is designed for the conditions the platform is about to create.
That question has been absent from every technology era the archive has documented. It is absent from the current AI cycle for the same reason: the providers cannot ask it because the honest answer sometimes requires a client to stop. The analyst firms cannot ask it because their research is funded by the vendors the answer might implicate. And procurement — the function closest to the implementation — does not have the organizational authority to make the answer matter.
The CIO does. The CFO does. And until they start asking it before the capital is committed rather than after the implementation fails, the failure rate will not move.
What RAM 2025™ surfaces that the current architecture cannot
The Mixture of Experts architecture inside every major AI-enabled platform your organization is currently running or evaluating is optimized toward one output: a confident, coherent recommendation from within its own training environment. It is not designed to challenge its own epistemological foundations. It is not designed to ask whether your governance architecture can act on what it is producing.
RAM 2025™ — the multimodel validation framework grounded in SR&ED-funded government research first implemented at Canada’s Department of National Defence in 1998 — operates on a different principle. It deploys genuinely independent models against the same problem in isolation and uses the divergence between them to find the exact boundary between what can be confidently claimed and what remains inference. It produces Stop signals. It produces Conditional signals. It identifies the governance gap before the capital is committed rather than after the consequences arrive.
The 1998 DND system improved delivery performance from 51% to 97.3% in three months because it was designed to be wrong and correct itself. The AI platforms CIOs are approving today are designed to be confident. The distance between those two design choices is not academic — it is the gap between a system that learns from being wrong and a system that hides its own uncertainty. That is the architectural generation the industry has not yet caught up to — and the archive has been documenting what happens when it does not for 28 years.
The reframe CIOs need to make before the next initiative
Stop asking: is this platform capable?
Start asking: is our organization designed to absorb what this platform is about to recommend?
The first question is answered by the vendor and rubber-stamped by the analyst firm the vendor funds. The second question has only one honest answer source: an independent organizational readiness diagnostic that is not optimized toward deployment and is not invested in any particular outcome except the truth.
That diagnostic exists. It has existed since 1998. And the CIOs who commission it before the next AI initiative will not be reading another post-mortem that assigns their capital allocation failure to the procurement function that was handed the implementation.
Technology is built to be right. RAM was built to get it right.
The difference between those two design philosophies is thirty years of failure reports — and the opportunity, right now, for the CIOs and CFOs who commission the diagnostic before the next initiative to be the first ones who do not generate another one.
The Procurement Insights archive contains 3,300+ published documents spanning 18 years of independently produced, timestamped procurement and supply chain research — zero vendor sponsorships, zero paid analyst relationships. Phase 0™ is the organizational readiness diagnostic that precedes all technology and supply chain commitments. RAM 2025™ is the multimodel validation framework that cross-validates all major Hansen Models™ assessments before publication.
If you are a CIO or CFO reading this and you felt that question land — you are the audience for this post. You already know the hot seat is real. You also know that if the diagnostic layer is still missing when the next initiative rolls out, the post-mortem will have your name on it. Not just the CPO’s.
Ready to close your Authority Gap?
If you have an active or upcoming AI or technology initiative, this is the moment to test whether your organization can act on what it will produce.
Book a 30-Minute Readiness Conversation with Jon Hansen — a preliminary diagnostic discussion to identify whether a Phase 0™ assessment is the right next step for your initiative. No sales pitch. Just an honest conversation about where your organization sits on the readiness spectrum.
Procurement Insights 
We Have Been Calling Them Procurement Failures. They Are Not. Here Is How CIOs and CFOs Can Own This Before the Next One.
Posted on March 27, 2026
0
For thirty years, the post-mortem has landed in the same place — and in the wrong office.
The ERP implementation underperforms. The eProcurement rollout stalls. The P2P platform goes live and the adoption numbers disappoint. The AI layer produces confident recommendations that do not convert into outcomes. And the narrative — in boardrooms, in analyst reports, in industry conference keynotes — assigns the failure to procurement.
Wrong function. Wrong diagnosis. Same result, every time.
We have been treating a capital allocation and governance failure as a procurement failure.
Who actually owns the failure
The CIO approved the architecture. The CFO approved the capital. The technology selection was made inside an evaluation framework designed by vendor-aligned analysts whose research was funded by the platforms being assessed. The implementation was handed to a procurement function that had no authority over the governance architecture it was being asked to operate within.
When the implementation failed — and across seven consecutive technology eras, it failed to meet expected outcomes between 60 and 85 percent of the time — consistently, across independent analyses and 28 years of archive documentation — procurement absorbed the accountability.
The function with the least authority inherited the consequences of decisions made by the functions with the most.
How do you sign off on a multimillion-dollar stack and not own it when it goes down the crapper?
That is what every practitioner sitting through a post-mortem has wanted to say out loud for thirty years.
That is not a procurement failure. That is a capital allocation failure. It is a governance architecture failure. It is an IT investment failure. And the CIOs and CFOs who have been nodding along at procurement conferences while the CPO presented the post-mortem have been watching the wrong function take responsibility for their decisions for three decades.
What the market is telling you right now
The procurement industry is contracting. Membership bodies are tightening focus. Community platforms are consolidating. Advisors who built their careers around procurement are quietly moving toward AI ventures and capital-backed platforms — toward the functions where transformation budget and decision authority actually sit.
This is not coincidence. This is the market correcting for a structural reality that has been true since 1995: procurement can diagnose the problem. Procurement cannot fund the solution. The mandate and the budget to fix structural failure sit with the CIO and the CFO — and they always have.
Deloitte’s CPO surveys have been documenting the downstream effects of this for years. More than 70% of CPOs report difficulty attracting and retaining talent capable of delivering strategic outcomes. The confidence gap between high-performing and low-performing procurement organizations is not primarily a training problem — it is an investment problem. Organizations that treat procurement as a strategic function develop their people accordingly. Organizations that treat procurement as a cost center hand their CPO a strategic mandate and a tactical budget.
The 2007 version of this problem was visible in the archive. The procurement profession’s strategic ambitions were outpacing executive willingness to invest in, empower, and retain the talent needed to produce strategic outcomes. Deloitte did not overturn that finding in 2023. They measured it.
The profession has been asking for a seat at the table for thirty years. The CIO and CFO have been handing them the chair and keeping the keys.
Why the current AI cycle is different — and more dangerous
Every previous technology era produced visible failure signals. A poorly configured ERP produced error messages. A poorly deployed P2P platform produced process bottlenecks practitioners could identify and report.
A poorly deployed AI procurement platform produces confident, fluent, coherent wrong answers — and the organization consuming those outputs has no native mechanism to distinguish them from correct ones. The danger is not that AI is mystical. It is that its mistakes are harder to see, arrive faster, and are easier for executives to rationalize. The MoE-based platforms that CIOs are currently approving and CFOs are currently funding are optimized to produce the most coherent answer from within their own data environment. They are not designed to tell you what they do not know. They are not designed to produce a Stop signal. And the analyst firms evaluating them — funded by the vendors being assessed — are not designed to produce one either.
The CIO who approved the architecture is accountable for what that architecture cannot do. The CFO who approved the capital is accountable for the ROI that architecture cannot deliver. Procurement will be handed the implementation. Procurement will absorb the post-mortem.
Unless something changes before the deployment begins.
The architectural question CIOs are not asking
The platforms are more capable than they have ever been. That is true and the CIOs who approved them are not wrong about what the technology can do in a controlled evaluation environment.
The question that is not being asked — by the CIO, by the CFO, by any vendor whose commercial model depends on the answer being proceed — is whether the organization’s governance architecture is designed to absorb what the platform is about to recommend.
Not whether the platform is capable. Not whether the data is clean. Not whether the integration is complete. Whether the human governance layer — the decision authority, the approval thresholds, the process integrity, the named individual who can act on a recommendation within the 30-day window that determines whether the response is a decision or a consequence — is designed for the conditions the platform is about to create.
That question has been absent from every technology era the archive has documented. It is absent from the current AI cycle for the same reason: the providers cannot ask it because the honest answer sometimes requires a client to stop. The analyst firms cannot ask it because their research is funded by the vendors the answer might implicate. And procurement — the function closest to the implementation — does not have the organizational authority to make the answer matter.
The CIO does. The CFO does. And until they start asking it before the capital is committed rather than after the implementation fails, the failure rate will not move.
What RAM 2025™ surfaces that the current architecture cannot
The Mixture of Experts architecture inside every major AI-enabled platform your organization is currently running or evaluating is optimized toward one output: a confident, coherent recommendation from within its own training environment. It is not designed to challenge its own epistemological foundations. It is not designed to ask whether your governance architecture can act on what it is producing.
RAM 2025™ — the multimodel validation framework grounded in SR&ED-funded government research first implemented at Canada’s Department of National Defence in 1998 — operates on a different principle. It deploys genuinely independent models against the same problem in isolation and uses the divergence between them to find the exact boundary between what can be confidently claimed and what remains inference. It produces Stop signals. It produces Conditional signals. It identifies the governance gap before the capital is committed rather than after the consequences arrive.
The 1998 DND system improved delivery performance from 51% to 97.3% in three months because it was designed to be wrong and correct itself. The AI platforms CIOs are approving today are designed to be confident. The distance between those two design choices is not academic — it is the gap between a system that learns from being wrong and a system that hides its own uncertainty. That is the architectural generation the industry has not yet caught up to — and the archive has been documenting what happens when it does not for 28 years.
The reframe CIOs need to make before the next initiative
Stop asking: is this platform capable?
Start asking: is our organization designed to absorb what this platform is about to recommend?
The first question is answered by the vendor and rubber-stamped by the analyst firm the vendor funds. The second question has only one honest answer source: an independent organizational readiness diagnostic that is not optimized toward deployment and is not invested in any particular outcome except the truth.
That diagnostic exists. It has existed since 1998. And the CIOs who commission it before the next AI initiative will not be reading another post-mortem that assigns their capital allocation failure to the procurement function that was handed the implementation.
Technology is built to be right. RAM was built to get it right.
The difference between those two design philosophies is thirty years of failure reports — and the opportunity, right now, for the CIOs and CFOs who commission the diagnostic before the next initiative to be the first ones who do not generate another one.
The Procurement Insights archive contains 3,300+ published documents spanning 18 years of independently produced, timestamped procurement and supply chain research — zero vendor sponsorships, zero paid analyst relationships. Phase 0™ is the organizational readiness diagnostic that precedes all technology and supply chain commitments. RAM 2025™ is the multimodel validation framework that cross-validates all major Hansen Models™ assessments before publication.
If you are a CIO or CFO reading this and you felt that question land — you are the audience for this post. You already know the hot seat is real. You also know that if the diagnostic layer is still missing when the next initiative rolls out, the post-mortem will have your name on it. Not just the CPO’s.
Ready to close your Authority Gap?
If you have an active or upcoming AI or technology initiative, this is the moment to test whether your organization can act on what it will produce.
Book a 30-Minute Readiness Conversation with Jon Hansen — a preliminary diagnostic discussion to identify whether a Phase 0™ assessment is the right next step for your initiative. No sales pitch. Just an honest conversation about where your organization sits on the readiness spectrum.
→ Book your 30-Minute Readiness Conversation
hansenprocurement.com
-30-
Share this:
Related