A $50-billion industry has spent decades trying to kill the spreadsheet. The spreadsheet is still here. That is not a failure of adoption — it is the physics showing through.
For twenty-five years, the spreadsheet has been the tool everyone agrees is wrong. It is error-prone, fragile, ungoverned, unauditable. Every enterprise platform since has been sold, in part, as the thing that would finally replace it. And by one widely cited estimate, after all of that, roughly 88% of organizations still run core business functions on it — against an ERP market that takes in something like $50 billion a year, much of it promising to make the spreadsheet obsolete.
The industry has a standard explanation for this, and the explanation is exactly backwards.
The standard read is that the spreadsheet survives despite being inferior — a failure of adoption, a matter of user habit, a problem to be solved by finally getting people onto the superior platform. On that view, persistence is a defect to be engineered away.
The physics reads it the other way. The spreadsheet survives because it imposes no operating model. It is the one tool that adapts to the organization’s actual conditions instead of forcing the organization to adapt to it. Its persistence is not the bug. It is the single clearest piece of everyday evidence for a claim I have been documenting across every technology era: the determining variable of an outcome is not the capability of the technology. It is the fit between the technology and the real conditions it enters.
A tool with no opinion
Every “superior” platform arrives with an opinion. An ERP, a planning suite, an agentic workflow — each encodes a model of how the work should happen: these fields, this sequence, these hand-offs, these roles. That embedded model is the platform’s great strength when it matches the organization’s real operating conditions, and its great liability when it does not. A strong opinion about how work should flow is an asset only when it happens to be the right opinion. When it is wrong, the platform does not bend. The organization is asked to bend instead.
A spreadsheet has no opinion. It is a blank surface that lets the user encode whatever the actual conditions are — the real order timing, the real exception, the real relationship between two numbers that the platform’s schema has no field for. It bends to the organization’s structure rather than overwriting it. Its famous “inefficiency” is the price of that adaptability — a price organizations keep paying because fit to real conditions is what actually determines whether the work gets done.
This is why the most telling pattern in the research is not avoidance but reversion. Finance teams leave Excel for a sophisticated platform — promising demos, real initial adoption — and then gradually return to the spreadsheet for the final reporting, the board decks, the scenario modeling. One account names it the “Excel paradox”: teams implement tools to escape spreadsheets, and keep coming back to spreadsheets because that is where the real work happens. Reversion is not weakness of will. It is the operating environment reasserting itself against an imposed model that did not fit it.
The test the industry keeps running and keeps misreading
There is a cleaner way to see what the spreadsheet proves. My standing claim is falsifiable in a single sentence: show me a case where the technology was the determining variable — superior capability winning on its own merits regardless of the conditions it entered. If capability alone governed outcomes, the superior platform would have displaced the inferior tool decades ago. That is what “superior” is supposed to mean.
It hasn’t. The better-resourced, better-architected, better-funded tool has been losing ground it should own to a spreadsheet, persistently, across every era. The market runs this experiment continuously and reads the result as a problem to be fixed. It is not a problem. It is the regularity, demonstrated by the most-used business tool on earth: capability is not the determinant. Fit to conditions is. The spreadsheet’s survival is not an anomaly in the data. It is the data.
What this is not
The point is not that spreadsheets are good and platforms are bad. The spreadsheet’s neutrality cuts both ways. The same blankness that lets it fit real conditions is exactly what makes it unable to enforce governance, scale control, or prevent the version chaos and error propagation that genuinely do plague it. The 88% figure is not a vindication of the spreadsheet; it is a measurement of how often the imposed model failed to fit.
So the honest claim is narrow and precise. The spreadsheet’s persistence does not prove the spreadsheet is right. It proves that adaptability to conditions beats embedded capability as the determinant of outcomes. Which is why the answer is never “ban the spreadsheet” and never “trust the spreadsheet.” It is to understand the operating conditions first, and then decide how much opinion the tool should carry — how much of the organization’s real structure it must bend to, and how much governance it must impose in return. The discipline is simply to sequence the diagnosis before the deployment.
The agentic edge
Trace the line forward and the stakes rise with it. A spreadsheet has no opinion; an ERP has a rigid one; the tool arriving next holds the strongest opinion yet. An agentic system does not merely encode an operating model — it acts on one, autonomously. The more capable and independent the tool, the stronger its embedded assumptions about how the work happens, and the more faithfully it will amplify a misalignment when those assumptions do not match the real environment. An agent that scales the wrong operating model does not produce a slightly worse spreadsheet. It produces a misalignment at machine speed.
The spreadsheet quietly persisted through five technology eras for one reason: it never asked the organization to become something it wasn’t. The tools now being sold to replace it are louder, more capable, and far more opinionated than anything before. The spreadsheet that refused to die is the warning label for the agent era — a twenty-five-year reminder, sitting on every desktop, that the tool which fits the conditions beats the tool that is merely superior, every single time the two are put to the test.
The names of the technologies change. The variable that decides the outcome does not.
Truth Is Believing. Accuracy Is Knowing.
Jon Hansen is the creator of Implementation Physics™, a research-based framework developed over nearly three decades to explain why technology initiatives succeed or fail regardless of the technology being deployed. His work spans six technology generations — from ERP through Agentic AI — and includes the Metaprise™ model first articulated in the late 1990s. His research forms the foundation for the Hansen Method™, Hansen Fit Score™ (HFS™), Phase 0™ Readiness Assessment, and the ARA™ RAM 2025™ multimodel verification architecture. He currently serves as a Board Member of the CIPS Americas Chapter.
-30-
Related
The Spreadsheet Test: Why the Inferior Tool Keeps Winning
Posted on June 22, 2026
0
A $50-billion industry has spent decades trying to kill the spreadsheet. The spreadsheet is still here. That is not a failure of adoption — it is the physics showing through.
For twenty-five years, the spreadsheet has been the tool everyone agrees is wrong. It is error-prone, fragile, ungoverned, unauditable. Every enterprise platform since has been sold, in part, as the thing that would finally replace it. And by one widely cited estimate, after all of that, roughly 88% of organizations still run core business functions on it — against an ERP market that takes in something like $50 billion a year, much of it promising to make the spreadsheet obsolete.
The industry has a standard explanation for this, and the explanation is exactly backwards.
The standard read is that the spreadsheet survives despite being inferior — a failure of adoption, a matter of user habit, a problem to be solved by finally getting people onto the superior platform. On that view, persistence is a defect to be engineered away.
The physics reads it the other way. The spreadsheet survives because it imposes no operating model. It is the one tool that adapts to the organization’s actual conditions instead of forcing the organization to adapt to it. Its persistence is not the bug. It is the single clearest piece of everyday evidence for a claim I have been documenting across every technology era: the determining variable of an outcome is not the capability of the technology. It is the fit between the technology and the real conditions it enters.
A tool with no opinion
Every “superior” platform arrives with an opinion. An ERP, a planning suite, an agentic workflow — each encodes a model of how the work should happen: these fields, this sequence, these hand-offs, these roles. That embedded model is the platform’s great strength when it matches the organization’s real operating conditions, and its great liability when it does not. A strong opinion about how work should flow is an asset only when it happens to be the right opinion. When it is wrong, the platform does not bend. The organization is asked to bend instead.
A spreadsheet has no opinion. It is a blank surface that lets the user encode whatever the actual conditions are — the real order timing, the real exception, the real relationship between two numbers that the platform’s schema has no field for. It bends to the organization’s structure rather than overwriting it. Its famous “inefficiency” is the price of that adaptability — a price organizations keep paying because fit to real conditions is what actually determines whether the work gets done.
This is why the most telling pattern in the research is not avoidance but reversion. Finance teams leave Excel for a sophisticated platform — promising demos, real initial adoption — and then gradually return to the spreadsheet for the final reporting, the board decks, the scenario modeling. One account names it the “Excel paradox”: teams implement tools to escape spreadsheets, and keep coming back to spreadsheets because that is where the real work happens. Reversion is not weakness of will. It is the operating environment reasserting itself against an imposed model that did not fit it.
The test the industry keeps running and keeps misreading
There is a cleaner way to see what the spreadsheet proves. My standing claim is falsifiable in a single sentence: show me a case where the technology was the determining variable — superior capability winning on its own merits regardless of the conditions it entered. If capability alone governed outcomes, the superior platform would have displaced the inferior tool decades ago. That is what “superior” is supposed to mean.
It hasn’t. The better-resourced, better-architected, better-funded tool has been losing ground it should own to a spreadsheet, persistently, across every era. The market runs this experiment continuously and reads the result as a problem to be fixed. It is not a problem. It is the regularity, demonstrated by the most-used business tool on earth: capability is not the determinant. Fit to conditions is. The spreadsheet’s survival is not an anomaly in the data. It is the data.
What this is not
The point is not that spreadsheets are good and platforms are bad. The spreadsheet’s neutrality cuts both ways. The same blankness that lets it fit real conditions is exactly what makes it unable to enforce governance, scale control, or prevent the version chaos and error propagation that genuinely do plague it. The 88% figure is not a vindication of the spreadsheet; it is a measurement of how often the imposed model failed to fit.
So the honest claim is narrow and precise. The spreadsheet’s persistence does not prove the spreadsheet is right. It proves that adaptability to conditions beats embedded capability as the determinant of outcomes. Which is why the answer is never “ban the spreadsheet” and never “trust the spreadsheet.” It is to understand the operating conditions first, and then decide how much opinion the tool should carry — how much of the organization’s real structure it must bend to, and how much governance it must impose in return. The discipline is simply to sequence the diagnosis before the deployment.
The agentic edge
Trace the line forward and the stakes rise with it. A spreadsheet has no opinion; an ERP has a rigid one; the tool arriving next holds the strongest opinion yet. An agentic system does not merely encode an operating model — it acts on one, autonomously. The more capable and independent the tool, the stronger its embedded assumptions about how the work happens, and the more faithfully it will amplify a misalignment when those assumptions do not match the real environment. An agent that scales the wrong operating model does not produce a slightly worse spreadsheet. It produces a misalignment at machine speed.
The spreadsheet quietly persisted through five technology eras for one reason: it never asked the organization to become something it wasn’t. The tools now being sold to replace it are louder, more capable, and far more opinionated than anything before. The spreadsheet that refused to die is the warning label for the agent era — a twenty-five-year reminder, sitting on every desktop, that the tool which fits the conditions beats the tool that is merely superior, every single time the two are put to the test.
The names of the technologies change. The variable that decides the outcome does not.
Truth Is Believing. Accuracy Is Knowing.
Jon Hansen is the creator of Implementation Physics™, a research-based framework developed over nearly three decades to explain why technology initiatives succeed or fail regardless of the technology being deployed. His work spans six technology generations — from ERP through Agentic AI — and includes the Metaprise™ model first articulated in the late 1990s. His research forms the foundation for the Hansen Method™, Hansen Fit Score™ (HFS™), Phase 0™ Readiness Assessment, and the ARA™ RAM 2025™ multimodel verification architecture. He currently serves as a Board Member of the CIPS Americas Chapter.
-30-
Share this:
Like this:
Related