Perspective
I help large organisations make technology change economically accountable and operationally reliable.
I connect investment decisions, organisational design, engineering evidence and AI-enabled delivery, so leaders can decide what to fund, what to change, what can safely scale, and whether the benefits actually arrive.
The interesting question is not whether an organisation is transforming. It is whether the change is working, and whether the organisation can act on what it learns.
Everything I work on comes back to one problem: how organisations make consequential changes under uncertainty, check their effects, and learn to decide better.
Three levels, usually seen apart
01
Enterprise economics and decisions
Where to invest, what to stop, which outcomes are worth pursuing, and who owns the benefits.
A strategy executive may understand this level but see little of engineering reality.
02
Organisational and delivery systems
How decisions, incentives, team structures, capabilities, constraints and measurement shape actual performance.
An organisational consultant may understand this level but lack the instruments to test whether an intervention works.
03
Engineering change and AI assurance
How AI changes software work, what must be verified, where effort and risk move, and how those effects can be measured.
An engineering leader may understand this level but have little influence over capital allocation and incentives.
The Accountable Change Model
Six stages, one loop
- 1
Investment intent
What change is worth making, and why?
Portfolio governanceThe next investment decision is also a capacity decision
- 2
System design
What structures, roles and incentives make it possible?
Giving benefits an ownerThe Art of Creating Self-Organizing Teams
- 3
Engineering execution
How do people and AI produce and verify change?
- 4
Operational evidence
What actually changed in flow, quality and capacity?
- 5
Economic outcomes
What benefit arrived, and who owns it?
From AI productivity to economic valueGiving benefits an owner
- 6
Investment feedback
What should we continue, redirect or stop?
The Predictability Loop is the method for stages three to five: how to trust a signal, investigate a constraint, test an intervention and make a defensible commitment.
This is a synthesis of my work, not a claim that all six stages have been implemented as one enterprise-wide framework.
When the evidence changes the decision
Connecting investment, organisational design and engineering evidence is worth the effort when it produces a different decision. Four situations show the difference.
01
A faster task can be the wrong investment
AI-assisted coding can reduce implementation effort while increasing the work needed for review and verification. Measured on coding speed alone, the investment looks successful; followed through the delivery system, the constraint may simply have moved.
The decision changes: Address the constraint that limits end-to-end delivery before expanding coding capacity on the strength of a productive-looking pilot. In one AI evaluation I led, that meant recommending changes to review and verification before wider expansion.
Following the time AI savesFrom AI productivity to economic value
02
A green project can be a poor investment
An initiative can meet every milestone while producing little evidence that its intended business outcome is becoming more likely. Delivery reporting says continue; investment reasoning asks whether the same people, time and money could produce more value elsewhere.
The decision changes: Continuing becomes an explicit investment choice instead of the automatic consequence of being on schedule. In one portfolio review, outcome evidence prompted senior leaders to redirect an initiative that was otherwise delivering to plan.
03
Released capacity is not automatically a financial return
Reducing the effort an activity needs creates capacity. On its own, that capacity does not reduce expenditure or improve customer outcomes.
The decision changes: A savings claim can't be approved because a task now takes less time. The next decision is what the capacity will be used for, which constraint prevents its productive use, and whether the intended benefit actually occurred.
Following the time AI savesFrom AI productivity to economic value
04
Shipping a product is not the same as realising its benefit
Delivery responsibility often ends at release, while the commercial or operational benefit depends on adoption, process changes and decisions made long after deployment. A completed delivery plan does not show that the investment succeeded.
The decision changes: Every significant investment carries a benefit hypothesis, a responsible owner and a post-release evidence review, and the result informs the next portfolio decision instead of disappearing into a completion report.
I don't treat delivery performance, organisational behaviour and investment economics as separate management conversations. I connect them at the point where a consequential decision has to be made.
The purpose is to establish what evidence is enough to continue, redirect, scale or stop an investment, and what remains uncertain.
One problem, several disciplines
That is why I write about financial measurement one week, organisational coaching the next, and AI verification after that. They are parts of the same problem.
Systems thinking with quantitative evidence
How delivery actually behaves, where its constraints are, and improvement told apart from noise.
Flow ObservatoryInvestment governance combined with delivery reality
Why organisations fund work, and why the benefits they expect fail to arrive.
Portfolio governanceOrganisational psychology combined with structural change
Changing decision rights through relationships, authority and behaviour as well as structure.
The benefit-ownership caseTurning ideas into instruments
Conceptual models turned into books, tools and toolkits others can use.
Research and toolsThe structural integrity of AI-assisted changeEmerging
What AI-assisted changes do to a codebase's structure. Presented as research, not as an established enterprise capability.
Coherence debt