Gartner is running a webinar on how to evaluate and select “Digital Twin of the Organization” (DTO) platforms, and the framing is exactly right: “Choosing a DTO platform is not just a technology decision. It is a strategic one.” I want to start by agreeing, because it is the most important sentence in the announcement. A digital twin of your organization is not a tool you buy. It is a decision about how you understand yourself.
So let me build on it — because “strategic, not technological” opens a door worth walking through.
The stack assumes something
Gartner’s “Building blocks of a digital twin of an organization.” Source: Gartner. Shown here for commentary and context.
A digital twin of the organization is a model — a living map of how the business works: its model, its operations, its performance, its actions. Gartner’s own framework stacks it cleanly, and the architecture is sound. Destination, Map, Performance, Situation, Value: a logical progression from business model down to business action.
But look closely at what every layer in that stack has in common. Each one represents the organization. None of them validates that the representation is correct. There is no Discovery layer, and no Validation layer. The stack quietly assumes you already understand the organization accurately — and then models that understanding with great sophistication.
The building block the DTO stack assumes but doesn’t include: Phase 0™ — validating the original before you model it. (Hansen Models™.)
That assumption is the whole game. Because a digital twin can only ever be as accurate as the assumptions embedded within it. If those assumptions are wrong, you don’t get a wrong answer you can catch. You get a flawless, high-resolution model of the wrong organization — confidently, at scale.
The question no model contained
Here is why that isn’t hypothetical.
In 1998, I was brought into a Department of National Defence engagement that was struggling. The reports were accurate. The process maps were accurate. The workflows were accurate. The technology was accurate. By every internal measure, the organization understood itself.
And yet no one had asked: what time of day do orders come in?
It wasn’t in any document or any system. But the answer — that orders clustered at a predictable time — reorganized everything downstream, and on-time delivery went from 51% to 97.3% in three months. I wrote about why this matters for AI here.
The point for digital twins is precise. That insight was not derived from process maps or ERP data or org charts. It came from a human noticing a behavioral pattern that no model was representing. A digital twin can model only what it already knows exists. The DND discovery was about finding something no model had captured yet — which is a different capability entirely, and it is almost a prerequisite to building a twin worth having. The organization’s own representation of itself was incomplete, and a twin built on it would have inherited the blind spot perfectly.
That is Implementation Physics™: it asks how the assumptions were validated before they became the twin. Not a criticism of digital twins — a question about what comes immediately before them.
There is a name for that “before.” I call it Phase 0™ — the discovery-and-validation work that establishes how the organization actually creates value, before any platform is chosen to model it. Phase 0™ is where you go looking for the question no one has asked yet. Skip it, and the twin faithfully digitizes whatever you already believed — blind spots included. Do it first, and the twin has a validated original to be true to.
Twenty years, and the same recurring problem
Which brings me to my second question: how long have digital twins been around, and how well have they actually worked?
The answer is instructive. The concept is not new. It traces to the early 2000s, with the term appearing in NASA and U.S. Air Force reports around 2012. Gartner named digital twins a Top 10 Strategic Technology Trend in 2017, and predicted half of large industrial companies would be using them by 2021. Even there — in the industrial, physical-asset world where twins are most mature — adoption ran behind the forecast (a 2019 Gartner survey put actual enterprise deployment in the single digits). The “of the organization” version is newer and less proven still.
But here is what should give any buyer pause. Across nearly two decades, the problems that keep surfacing are remarkably consistent: data quality, model fidelity, synchronization with reality, unclear adoption, weak governance, and the struggle to keep the twin aligned with an organization that keeps changing.
Read that list again. Not one of those is a technology limitation. Every one of them is an operating-logic and human problem. The twin isn’t failing because the simulation is weak. It’s failing at the seams where the model meets the messy, moving, human reality it was supposed to represent.
This is Invariant Physics™
Step back far enough and a pattern appears that is older than digital twins.
Across every technology wave — ERP, cloud, digital twins, AI, agentic systems — the limiting factor keeps returning to the same place: trust, decision rights, human judgment, operating discipline, organizational readiness. The technology changes completely each era. The human constraint changes form — but the underlying limiting function stays remarkably constant. That is Invariant Physics™.
A digital twin promises to be a living model. But it is only ever as current as the mapping that feeds it — and “synchronization with reality” is on the industry’s own list of recurring failures for a reason. A model of a moving organization, captured once and assumed, becomes an increasingly confident portrait of a company that no longer exists. Keeping it true isn’t a platform feature. It’s governance, exercised continuously.
Validate the original
So evaluate the platforms. Gartner is right that it is a strategic decision, and the good frameworks are genuinely useful. I’d only add the question that sits one layer beneath the whole exercise:
Before you build the twin, how do you know you are twinning the right organization?
The quality of every digital twin is bounded by the quality of the organizational understanding that existed before the twin was ever constructed. That understanding is not produced by the platform. It is produced by the human discipline of validating how the organization actually creates value — the discipline that knows to ask what time of day the orders come in.
The technology changes every era. That constraint never does.
Truth Is Believing. Accuracy Is Knowing. Outcome Is Proof.™
-30-
Related
Before You Build the Twin, Validate the Original
Posted on July 8, 2026
0
Gartner is running a webinar on how to evaluate and select “Digital Twin of the Organization” (DTO) platforms, and the framing is exactly right: “Choosing a DTO platform is not just a technology decision. It is a strategic one.” I want to start by agreeing, because it is the most important sentence in the announcement. A digital twin of your organization is not a tool you buy. It is a decision about how you understand yourself.
So let me build on it — because “strategic, not technological” opens a door worth walking through.
The stack assumes something
Gartner’s “Building blocks of a digital twin of an organization.” Source: Gartner. Shown here for commentary and context.
A digital twin of the organization is a model — a living map of how the business works: its model, its operations, its performance, its actions. Gartner’s own framework stacks it cleanly, and the architecture is sound. Destination, Map, Performance, Situation, Value: a logical progression from business model down to business action.
But look closely at what every layer in that stack has in common. Each one represents the organization. None of them validates that the representation is correct. There is no Discovery layer, and no Validation layer. The stack quietly assumes you already understand the organization accurately — and then models that understanding with great sophistication.
The building block the DTO stack assumes but doesn’t include: Phase 0™ — validating the original before you model it. (Hansen Models™.)
That assumption is the whole game. Because a digital twin can only ever be as accurate as the assumptions embedded within it. If those assumptions are wrong, you don’t get a wrong answer you can catch. You get a flawless, high-resolution model of the wrong organization — confidently, at scale.
The question no model contained
Here is why that isn’t hypothetical.
In 1998, I was brought into a Department of National Defence engagement that was struggling. The reports were accurate. The process maps were accurate. The workflows were accurate. The technology was accurate. By every internal measure, the organization understood itself.
And yet no one had asked: what time of day do orders come in?
It wasn’t in any document or any system. But the answer — that orders clustered at a predictable time — reorganized everything downstream, and on-time delivery went from 51% to 97.3% in three months. I wrote about why this matters for AI here.
The point for digital twins is precise. That insight was not derived from process maps or ERP data or org charts. It came from a human noticing a behavioral pattern that no model was representing. A digital twin can model only what it already knows exists. The DND discovery was about finding something no model had captured yet — which is a different capability entirely, and it is almost a prerequisite to building a twin worth having. The organization’s own representation of itself was incomplete, and a twin built on it would have inherited the blind spot perfectly.
That is Implementation Physics™: it asks how the assumptions were validated before they became the twin. Not a criticism of digital twins — a question about what comes immediately before them.
There is a name for that “before.” I call it Phase 0™ — the discovery-and-validation work that establishes how the organization actually creates value, before any platform is chosen to model it. Phase 0™ is where you go looking for the question no one has asked yet. Skip it, and the twin faithfully digitizes whatever you already believed — blind spots included. Do it first, and the twin has a validated original to be true to.
Twenty years, and the same recurring problem
Which brings me to my second question: how long have digital twins been around, and how well have they actually worked?
The answer is instructive. The concept is not new. It traces to the early 2000s, with the term appearing in NASA and U.S. Air Force reports around 2012. Gartner named digital twins a Top 10 Strategic Technology Trend in 2017, and predicted half of large industrial companies would be using them by 2021. Even there — in the industrial, physical-asset world where twins are most mature — adoption ran behind the forecast (a 2019 Gartner survey put actual enterprise deployment in the single digits). The “of the organization” version is newer and less proven still.
But here is what should give any buyer pause. Across nearly two decades, the problems that keep surfacing are remarkably consistent: data quality, model fidelity, synchronization with reality, unclear adoption, weak governance, and the struggle to keep the twin aligned with an organization that keeps changing.
Read that list again. Not one of those is a technology limitation. Every one of them is an operating-logic and human problem. The twin isn’t failing because the simulation is weak. It’s failing at the seams where the model meets the messy, moving, human reality it was supposed to represent.
This is Invariant Physics™
Step back far enough and a pattern appears that is older than digital twins.
Across every technology wave — ERP, cloud, digital twins, AI, agentic systems — the limiting factor keeps returning to the same place: trust, decision rights, human judgment, operating discipline, organizational readiness. The technology changes completely each era. The human constraint changes form — but the underlying limiting function stays remarkably constant. That is Invariant Physics™.
A digital twin promises to be a living model. But it is only ever as current as the mapping that feeds it — and “synchronization with reality” is on the industry’s own list of recurring failures for a reason. A model of a moving organization, captured once and assumed, becomes an increasingly confident portrait of a company that no longer exists. Keeping it true isn’t a platform feature. It’s governance, exercised continuously.
Validate the original
So evaluate the platforms. Gartner is right that it is a strategic decision, and the good frameworks are genuinely useful. I’d only add the question that sits one layer beneath the whole exercise:
Before you build the twin, how do you know you are twinning the right organization?
The quality of every digital twin is bounded by the quality of the organizational understanding that existed before the twin was ever constructed. That understanding is not produced by the platform. It is produced by the human discipline of validating how the organization actually creates value — the discipline that knows to ask what time of day the orders come in.
The technology changes every era. That constraint never does.
Truth Is Believing. Accuracy Is Knowing. Outcome Is Proof.™
-30-
Share this:
Like this:
Related