For forty years, the major consulting firms have each published their own model of why transformation succeeds. Strip away the acronyms and they all say the same thing — and it isn’t technology. So why hasn’t the failure rate moved?
This blog carries the following words as a badge of honor:
Truth Is Believing. Accuracy Is Knowing.
A note on method before the argument, because the argument depends on it. Every framework named below was checked against its original, dated, published source — not recalled, not paraphrased from memory. McKinsey’s 7-S to its June 1980 Business Horizons article; BCG’s DICE and Bain’s RAPID to their Harvard Business Review publications; the rest to the record. Where a figure is contested, I say so. And the firms named here are invited to challenge any attribution — if I have a date, an author, or a claim wrong, I will correct it in public. The point of this piece is not to be right about these firms. It is to get it right.
Here is what the record shows.
The same diagnosis, four times over
In 1980, three McKinsey consultants — Robert Waterman, Tom Peters, and Julien Phillips — published “Structure Is Not Organization” and introduced the 7-S Framework. Its entire argument was that the “soft” elements — shared values, skills, staff, style — matter as much as the “hard” ones, and that performance comes from their alignment, not from structure alone.
In 1996, Harvard’s John Kotter published Leading Change and its eight steps, which became the canonical change-management model: success is led through people.
In 2005, BCG’s Harold Sirkin, Perry Keenan, and Alan Jackson published DICE in Harvard Business Review — Duration, Integrity, Commitment, Effort — built on the study of more than 200 change initiatives. The factors that predict success are executional and human.
In 2006, Bain’s Paul Rogers and Marcia Blenko published RAPID in Harvard Business Review — Recommend, Agree, Perform, Input, Decide. Transformation turns on decision rights and accountability.
And in 2017, Deloitte’s digital transformation and maturity frameworks landed in the same place: strategy, culture, and people, alongside the technology.
Different firms. Different decades. Different acronyms. One diagnosis: the determinant of success is the organization — its people, its processes, its decision rights, its culture — not the technology.
They were right. Every one of them. That is not the criticism. That is the setup.
Then why hasn’t anything changed?
Because for the same forty years, the failure rate has not moved.
The figure most often cited is “70% of transformations fail” — and in the interest of accuracy, that specific number is itself unsubstantiated: a 2011 study in the Journal of Change Management by Mark Hughes found no reliable empirical evidence for it and traced it to a 1993 book. But every credible measure since lands in the same grim territory. McKinsey’s own 2021 study, “Losing From Day One,” put it at 69%. A Bain study found only 12% of transformations fully succeeded, with 75% delivering mediocre results. MIT’s 2025 study found 95% of GenAI pilots produced no measurable return.
Four decades. The diagnosis was consistent and correct. The result didn’t improve. That is the puzzle the rest of this piece is about.
Where the firms went off road — and why
Here is the part I want to be precise about, because it is easy to get wrong and unfair if you do.
The firms did not omit people and process. The opposite — they named them, repeatedly, for forty years. The error was not omission. It was sequence.
Everyone in this field talks about “people, process, and technology.” I have talked about people, process, and technology for my entire career. But notice the word in the middle: and. “People, process, and technology” makes the three co-equal and simultaneous — a list of things to manage in parallel, a balanced trinity to optimize all at once. That framing is the wrong turn. Because they are not co-equal, and they are not simultaneous. People and process are not items on the list alongside technology. They are the preconditions for it. The correct formulation was never and. It was then: people and process first — technology last.
That is not a list. It is a dependency. And the firms, almost without exception, flattened the dependency into a list.
Why? Three structural reasons — none of them a matter of competence or good faith:
Commercial gravity. Technology is the sellable, scalable, licensable, billable deliverable. Substrate alignment is bespoke, points the client inward rather than toward a purchase, and cannot be turned into a product. So even when the framework says people and process come first, the business model leads with the technology, bolting on change management afterward. The firm’s incentives pull against its own diagnosis.
The legibility trap. A capability ladder generalizes — five clean levels, the same diagram for every client, reusable and teachable. Substrate alignment is the opposite: organization-specific, resistant to templating, impossible to productize. So frameworks drift toward what is legible and repeatable, and the substrate becomes a footnote — present in the words, absent from the sequence.
The trinity itself. Once “people, process, technology” became a balanced triad, the order disappeared. A checklist says do all three. A sequence says do these two before that one, or the third cannot land. The firms optimized the trinity. The dependency got lost in the conjunction.
So the firms did not go off road by ignoring the substrate. They went off road by de-sequencing it — naming the right variables, then reorganizing them from an order into an inventory, pulled by what sells and what scales.
I sequenced it in 2007 — in public, in print
This is not hindsight. On July 4, 2007, I published “Dangerous Supply Chain Myths (Part 7),” laying out a high-level three-step method in exactly this order:
Step 1 — understand the characteristics of your spend and what impacts the outcomes you want to achieve. Step 2 — identify and align the process with all agent or stakeholder input — even the agents you do not think are directly involved, but usually are — before evaluating any technology. Step 3 — then introduce technology, to accelerate the process rather than define it. In my own words at the time: “the technology was the final step in the process… it adapted to the real-world processes of the client.”
That sequencing came from a 1998 engagement — research partly funded by the Government of Canada’s SR&ED program — in which technology was introduced only after the spend analysis and process alignment were complete. The result: a public-sector organization realized roughly 23% cost-of-goods savings sustained over several years, while the number of buyers required to manage the contract fell from 23 to three, with delivery performance and quality improving. The technology was primitive by today’s standard. It worked because it was last, not first.
That is the road the frameworks did not take.
The fix is not a fifth framework
Here is what I am not saying. I am not offering a better acronym to set alongside 7-S, DICE, and RAPID. The diagnosis does not need improving — it has been correct for forty years.
What is missing is the enforcement of sequence. Phase 0™ is not a fifth framework. It is the precondition every one of these frameworks assumed and none of them enforced: align the substrate — people, process, operating conditions — before the technology decision, not as a parallel workstream beside it. Substrate first. Technology last. The thing the conjunction “and” quietly erased, put back as an order.
Today’s takeaway: The reason transformation has failed at roughly the same rate for four decades, through every technology era, is not that the firms named the wrong variable. They named the right one, four times over. It is that naming a variable was never the same as sequencing it first. Fix the sequence, and the technology — whichever era’s technology it happens to be — finally does what it promised.
Jon Hansen is the founder of Hansen Models™ and publisher of Procurement Insights, an independent research practice operating on a nineteen-year archive with zero vendor sponsorships. Framework attributions in this piece are drawn from original, dated sources; the firms named are invited to challenge any of them.
They All Named the Right Variable. None of Them Changed the Result.
Posted on June 6, 2026
0
For forty years, the major consulting firms have each published their own model of why transformation succeeds. Strip away the acronyms and they all say the same thing — and it isn’t technology. So why hasn’t the failure rate moved?
This blog carries the following words as a badge of honor:
Truth Is Believing. Accuracy Is Knowing.
A note on method before the argument, because the argument depends on it. Every framework named below was checked against its original, dated, published source — not recalled, not paraphrased from memory. McKinsey’s 7-S to its June 1980 Business Horizons article; BCG’s DICE and Bain’s RAPID to their Harvard Business Review publications; the rest to the record. Where a figure is contested, I say so. And the firms named here are invited to challenge any attribution — if I have a date, an author, or a claim wrong, I will correct it in public. The point of this piece is not to be right about these firms. It is to get it right.
Here is what the record shows.
The same diagnosis, four times over
In 1980, three McKinsey consultants — Robert Waterman, Tom Peters, and Julien Phillips — published “Structure Is Not Organization” and introduced the 7-S Framework. Its entire argument was that the “soft” elements — shared values, skills, staff, style — matter as much as the “hard” ones, and that performance comes from their alignment, not from structure alone.
In 1996, Harvard’s John Kotter published Leading Change and its eight steps, which became the canonical change-management model: success is led through people.
In 2005, BCG’s Harold Sirkin, Perry Keenan, and Alan Jackson published DICE in Harvard Business Review — Duration, Integrity, Commitment, Effort — built on the study of more than 200 change initiatives. The factors that predict success are executional and human.
In 2006, Bain’s Paul Rogers and Marcia Blenko published RAPID in Harvard Business Review — Recommend, Agree, Perform, Input, Decide. Transformation turns on decision rights and accountability.
And in 2017, Deloitte’s digital transformation and maturity frameworks landed in the same place: strategy, culture, and people, alongside the technology.
Different firms. Different decades. Different acronyms. One diagnosis: the determinant of success is the organization — its people, its processes, its decision rights, its culture — not the technology.
They were right. Every one of them. That is not the criticism. That is the setup.
Then why hasn’t anything changed?
Because for the same forty years, the failure rate has not moved.
The figure most often cited is “70% of transformations fail” — and in the interest of accuracy, that specific number is itself unsubstantiated: a 2011 study in the Journal of Change Management by Mark Hughes found no reliable empirical evidence for it and traced it to a 1993 book. But every credible measure since lands in the same grim territory. McKinsey’s own 2021 study, “Losing From Day One,” put it at 69%. A Bain study found only 12% of transformations fully succeeded, with 75% delivering mediocre results. MIT’s 2025 study found 95% of GenAI pilots produced no measurable return.
Four decades. The diagnosis was consistent and correct. The result didn’t improve. That is the puzzle the rest of this piece is about.
Where the firms went off road — and why
Here is the part I want to be precise about, because it is easy to get wrong and unfair if you do.
The firms did not omit people and process. The opposite — they named them, repeatedly, for forty years. The error was not omission. It was sequence.
Everyone in this field talks about “people, process, and technology.” I have talked about people, process, and technology for my entire career. But notice the word in the middle: and. “People, process, and technology” makes the three co-equal and simultaneous — a list of things to manage in parallel, a balanced trinity to optimize all at once. That framing is the wrong turn. Because they are not co-equal, and they are not simultaneous. People and process are not items on the list alongside technology. They are the preconditions for it. The correct formulation was never and. It was then: people and process first — technology last.
That is not a list. It is a dependency. And the firms, almost without exception, flattened the dependency into a list.
Why? Three structural reasons — none of them a matter of competence or good faith:
Commercial gravity. Technology is the sellable, scalable, licensable, billable deliverable. Substrate alignment is bespoke, points the client inward rather than toward a purchase, and cannot be turned into a product. So even when the framework says people and process come first, the business model leads with the technology, bolting on change management afterward. The firm’s incentives pull against its own diagnosis.
The legibility trap. A capability ladder generalizes — five clean levels, the same diagram for every client, reusable and teachable. Substrate alignment is the opposite: organization-specific, resistant to templating, impossible to productize. So frameworks drift toward what is legible and repeatable, and the substrate becomes a footnote — present in the words, absent from the sequence.
The trinity itself. Once “people, process, technology” became a balanced triad, the order disappeared. A checklist says do all three. A sequence says do these two before that one, or the third cannot land. The firms optimized the trinity. The dependency got lost in the conjunction.
So the firms did not go off road by ignoring the substrate. They went off road by de-sequencing it — naming the right variables, then reorganizing them from an order into an inventory, pulled by what sells and what scales.
I sequenced it in 2007 — in public, in print
This is not hindsight. On July 4, 2007, I published “Dangerous Supply Chain Myths (Part 7),” laying out a high-level three-step method in exactly this order:
Step 1 — understand the characteristics of your spend and what impacts the outcomes you want to achieve. Step 2 — identify and align the process with all agent or stakeholder input — even the agents you do not think are directly involved, but usually are — before evaluating any technology. Step 3 — then introduce technology, to accelerate the process rather than define it. In my own words at the time: “the technology was the final step in the process… it adapted to the real-world processes of the client.”
That sequencing came from a 1998 engagement — research partly funded by the Government of Canada’s SR&ED program — in which technology was introduced only after the spend analysis and process alignment were complete. The result: a public-sector organization realized roughly 23% cost-of-goods savings sustained over several years, while the number of buyers required to manage the contract fell from 23 to three, with delivery performance and quality improving. The technology was primitive by today’s standard. It worked because it was last, not first.
That is the road the frameworks did not take.
The fix is not a fifth framework
Here is what I am not saying. I am not offering a better acronym to set alongside 7-S, DICE, and RAPID. The diagnosis does not need improving — it has been correct for forty years.
What is missing is the enforcement of sequence. Phase 0™ is not a fifth framework. It is the precondition every one of these frameworks assumed and none of them enforced: align the substrate — people, process, operating conditions — before the technology decision, not as a parallel workstream beside it. Substrate first. Technology last. The thing the conjunction “and” quietly erased, put back as an order.
Today’s takeaway: The reason transformation has failed at roughly the same rate for four decades, through every technology era, is not that the firms named the wrong variable. They named the right one, four times over. It is that naming a variable was never the same as sequencing it first. Fix the sequence, and the technology — whichever era’s technology it happens to be — finally does what it promised.
Jon Hansen is the founder of Hansen Models™ and publisher of Procurement Insights, an independent research practice operating on a nineteen-year archive with zero vendor sponsorships. Framework attributions in this piece are drawn from original, dated sources; the firms named are invited to challenge any of them.
-30-
Share this:
Like this:
Related