The fourth wave of transformation
Each wave reached one level deeper into the work. Reasoning is the last level.
Xavier Vallée, Founder, Sapionic 5 min read
I have spent close to thirty years inside transformation programmes. Re-engineering, Lean, digital. Each one promised a step change. Each one delivered something real. And each one stopped at the same place.
The decision.
Three waves, three levels
Business process re-engineering, the 1990s. Michael Hammer's 1990 Harvard Business Review article was called "Reengineering Work: Don't Automate, Obliterate". The instinct was right: stop paving over a bad process and redraw it. Re-engineering changed the shape of the work. But in too many companies it became a label for cutting heads, and the people who had to run the new process were an afterthought.
Lean and Six Sigma, the 1990s to the 2000s. Toyota, then Motorola and GE, taught a generation of managers to remove waste and measure variation. Lean and Six Sigma improved the quality of the steps. They could measure a defect to the decimal point. What they could not touch was judgement: two experienced people looking at the same case and deciding differently. That stayed out of scope.
Digital transformation, the 2000s to the 2010s. Forms went online. Paper disappeared. Systems started talking to each other. Digital removed the manual handling between the steps. But the decisions stayed exactly where they were: in people's heads, in email threads, in meetings. Robert Solow's line from 1987 applied all over again:
"You can see the computer age everywhere but in the productivity statistics."
The boundary every wave hit
Look at what the three waves had in common. Each reached one level deeper into the work: the shape of the process, the quality of the steps, the handling between them. And each stopped at the same boundary.
For thirty years, the only decisions you could automate were the ones you could reduce to a rule. Anything resting on a document, a precedent, a piece of context or an experienced opinion stayed with a person.
So we built the work around those people. The queues, the hand-offs, the approvals, the status calls. The research pack is assembled so a decision can be made. The deck is written so a decision can be made. The meeting is held so a decision can be made. Workflows are scaffolding around decisions.
And that is where the cost, the delay and the variance sit.
Why now
That constraint has just been removed.
Models can now read a document, weigh context, apply precedent and explain their reasoning. So the decisions that rest on judgement can be codified and made by a system, consistently, hundreds of times a day, with your experts handling the exceptions.
This is the fourth wave: decision-centric AI transformation. It does not optimise the process, the steps or the hand-offs. It optimises the reasoning itself. Reasoning is the last level.
1990s
Business Process Re-engineering
Redrew the process.
Optimised: the shape of the work
1990s to 2000s
Lean & Six Sigma
Removed the waste and the variation.
Optimised: the quality of the steps
2000s to 2010s
Digital transformation
Digitised the steps and the hand-offs.
Optimised: the manual handling
Now
Decision-Centric AI Transformation
Codifies the reasoning inside them.
Optimised: the reasoning itself
Each wave reached one level deeper into the work. Reasoning is the last level.
The economics move when the decision moves. A process that took a trained expert three to five hours by hand now runs in 230.7 seconds. Not a faster old process. A different kind of system.
Most AI programmes are running the third wave again
Here is the trap. Most companies are running AI as if it were the third wave. Roll out a tool. Automate some tasks. Count the hours saved.
Rolling out Copilot and expecting the benefits of AI is like rolling out Outlook and expecting the benefits of digital transformation.
The documents arrive faster. The decisions do not move. And economists are already reaching for Solow's line again.
What the fourth wave must learn from the first three
Each earlier wave left a lesson. The fourth wave needs all of them.
From re-engineering: do not automate the old process. Redesign the work around the decision. Hammer was right about that. An agent on a broken process is a faster broken process.
From Six Sigma: measure the variance. The new defect is decision variance: the same case, decided differently depending on who picks it up. It is the cost nobody measures, and it is often the largest. Baseline it before you build, or you will never prove the benefit.
From re-engineering, again: lead the change. This wave is more personal than the others. Digital changed the tools people worked with. This changes the reasoning they were valued for. That has to be led, not announced.
From digital: make the value reach the accounts. Efficiency that never reaches the P&L is not value. If you cannot trace a saving to a line in the accounts, assume it is not there.
What it means for leaders and investors
For a CEO, the operating model changes. A decision becomes a system in its own right, with an owner, a record of every choice it makes, and a marginal cost close to zero. Volume stops being tied to headcount. Your best expert's judgement becomes available on every case, not only the ones they have time for.
For an investor, this belongs in the value creation plan, not the IT budget. Codified decisions are an asset. They do not leave when people do, they get cheaper with every reuse, and across a portfolio they compound.
The first three waves changed how the work moves. The fourth changes how the business decides.
Treat it like the third and you get faster documents. Treat it as the fourth and the economics move.