There was a moment when navigating without GPS stopped being a skill and started being a risk.
Supply chain planning may have reached a similar tipping point. Digital twins and AI are rapidly becoming core planning infrastructure — not because they are exciting and new, but because the world supply chains now operate in “leaves little alternative.”
That last phrase should give you pause.
“Leaves little alternative” is the exact narrative that precedes most transformation failures.
I call it the Torrent Argument — the belief that “we have no choice” becomes a reliable predictor of failure, not because the technology is wrong, but because urgency replaces governance.
The Divergence Point
Digital twins represent an equation-based approach to supply chain management:
Variables are fixed inputs
The system is modeled mathematically
Optimizes for efficiency
Assumes rational actors
“The model says do X.”
Phase 0™ represents an agent-based approach:
Variables emerge from behavior
The system is mapped relationally
Optimizes for absorptive capacity
Accounts for human judgment, politics, and resistance
“Can your organization actually do X?”
What Digital Twins Cannot Encode
A digital twin models:
Supplier lead times
Inventory levels
Demand patterns
Logistics constraints
A digital twin cannot model:
Who has decision rights when the model recommends action
Whether stakeholders will trust the output
How exceptions are governed
Who’s accountable when the twin is wrong
The political dynamics that override “optimal” recommendations
The twin assumes the organization will act on what it shows. Phase 0™ asks whether the organization can act.
The Missing Step
The digital twin paradigm assumes:
Model reality → Recommend action → Organization acts → Success
Phase 0™ inserts the missing step:
Model reality → Assess organizational readiness to govern what the model reveals → If ready, act → Success. If not, the twin just accelerates dysfunction.
Equation-Based vs. Agent-Based
Equation-Based (Digital Twin)
Agent-Based (Phase 0™)
Assumes variables are known
Surfaces variables that aren’t
Models the supply chain
Maps the organization
Optimizes for efficiency
Optimizes for absorptive capacity
“The model says do X”
“Can your organization do X?”
Built on assumptions
Built on experience and constraints
Certainty theater
Decision collaboration
The 1998 Origin Point
In early agent-based work in 1998, systems tracked historic and real-time variables and presented them to the buyer. The buyer then decided the weighting:
Is this time-sensitive?
Or cost-sensitive?
Does risk outweigh savings?
Does resilience outweigh efficiency?
The technology handled complexity. The human handled meaning.
That wasn’t automation. It was decision collaboration.
Twenty-seven years later, digital twins are exponentially more powerful. But the fundamental truth hasn’t changed:
AI does not replace judgment. It stimulates dialogue.
Digital twins that operate as oracles — “the model will show you the answer” — are repeating the same mistake every failed technology wave has made since ERP.
The Torrent Connection
When someone says “the world leaves little alternative,” they’re invoking the torrent narrative.
The urgency feels real. The pressure is real. But the choice is still real, too.
Sequencing is still a choice
Scope is still a choice
Governance is still a choice
Readiness is still a choice
What changes is tolerance for ambiguity — not the laws of transformation.
The Question No Model Can Ask
At the Department of National Defence, the breakthrough came from a simple question: “What time of day do your orders come in?”
That question — which cost nothing to ask — revealed a pattern no simulation had ever modeled. It surfaced a constraint that had been invisible for years. And it changed everything.
At the New York City Transit Authority, that same question would have been irrelevant. The constraint was different. But the methodology surfaced it anyway — because agent-based approaches don’t assume the variables are known. They discover them.
A Digital Twin can only model what someone thought to include. It cannot ask the question no one knew to ask.
It’s just as easy to ask “What time of day do your orders come in?” as it is to build a Digital Twin. Easier, actually. And often more effective.
Because the question surfaces the constraint the twin would never have modeled — since no one knew to include it.
That’s the difference between assuming variables are known and surfacing the ones that aren’t.
The Bottom Line
A digital twin without governance is an engine without a steering wheel.
It will model the supply chain beautifully. It will recommend actions optimally. And it will scale dysfunction faster than any technology before it — because the organization was never assessed for its capacity to act on what the twin reveals.
The question isn’t whether digital twins are necessary.
The question is whether your organization is ready to govern them.
That readiness doesn’t emerge from the technology. It has to be designed — before deployment, not after.
Jon Hansen is the creator of The Hansen Method® and founder of Hansen Models™, helping organizations prevent the 80% implementation failure rate through Phase 0™ readiness assessment.
Procurement Insights
Digital Twins and the Torrent Trap: Why Modeling the Supply Chain Isn’t the Same as Governing It
Posted on January 27, 2026
0
By Jon W. Hansen | Procurement Insights
There was a moment when navigating without GPS stopped being a skill and started being a risk.
Supply chain planning may have reached a similar tipping point. Digital twins and AI are rapidly becoming core planning infrastructure — not because they are exciting and new, but because the world supply chains now operate in “leaves little alternative.”
That last phrase should give you pause.
“Leaves little alternative” is the exact narrative that precedes most transformation failures.
I call it the Torrent Argument — the belief that “we have no choice” becomes a reliable predictor of failure, not because the technology is wrong, but because urgency replaces governance.
The Divergence Point
Digital twins represent an equation-based approach to supply chain management:
Phase 0™ represents an agent-based approach:
What Digital Twins Cannot Encode
A digital twin models:
A digital twin cannot model:
The twin assumes the organization will act on what it shows. Phase 0™ asks whether the organization can act.
The Missing Step
The digital twin paradigm assumes:
Phase 0™ inserts the missing step:
Equation-Based vs. Agent-Based
The 1998 Origin Point
In early agent-based work in 1998, systems tracked historic and real-time variables and presented them to the buyer. The buyer then decided the weighting:
The technology handled complexity. The human handled meaning.
That wasn’t automation. It was decision collaboration.
Twenty-seven years later, digital twins are exponentially more powerful. But the fundamental truth hasn’t changed:
AI does not replace judgment. It stimulates dialogue.
Digital twins that operate as oracles — “the model will show you the answer” — are repeating the same mistake every failed technology wave has made since ERP.
The Torrent Connection
When someone says “the world leaves little alternative,” they’re invoking the torrent narrative.
The urgency feels real. The pressure is real. But the choice is still real, too.
What changes is tolerance for ambiguity — not the laws of transformation.
The Question No Model Can Ask
At the Department of National Defence, the breakthrough came from a simple question: “What time of day do your orders come in?”
That question — which cost nothing to ask — revealed a pattern no simulation had ever modeled. It surfaced a constraint that had been invisible for years. And it changed everything.
At the New York City Transit Authority, that same question would have been irrelevant. The constraint was different. But the methodology surfaced it anyway — because agent-based approaches don’t assume the variables are known. They discover them.
A Digital Twin can only model what someone thought to include. It cannot ask the question no one knew to ask.
It’s just as easy to ask “What time of day do your orders come in?” as it is to build a Digital Twin. Easier, actually. And often more effective.
Because the question surfaces the constraint the twin would never have modeled — since no one knew to include it.
That’s the difference between assuming variables are known and surfacing the ones that aren’t.
The Bottom Line
A digital twin without governance is an engine without a steering wheel.
It will model the supply chain beautifully. It will recommend actions optimally. And it will scale dysfunction faster than any technology before it — because the organization was never assessed for its capacity to act on what the twin reveals.
The question isn’t whether digital twins are necessary.
The question is whether your organization is ready to govern them.
That readiness doesn’t emerge from the technology. It has to be designed — before deployment, not after.
Jon Hansen is the creator of The Hansen Method® and founder of Hansen Models™, helping organizations prevent the 80% implementation failure rate through Phase 0™ readiness assessment.
-30-
Share this:
Related