Cases. Patterns. Structural failures.
Each pattern demonstrates the framework applied to a specific organisational failure mode.
These are not standalone observations — they are diagnostic applications showing how decision authority and incentives create rational failure.
When productivity improvement is not institutionally protected
Improving the work isolated people instead of changing the system.
A new graduate joined a team responsible for monthly reporting and lead generation.
In the first month, he automated his reporting work.
Quality improved.
Turnaround time shortened.
By the second month, the work was finished early.
Overtime stopped.
With time freed, he began producing analysis on top of reporting.
Insights replaced repetition.
Around him, others doing the same work were still working long hours.
One colleague asked for help.
The work was automated.
Overtime dropped.
Encouraged, he shared the idea more broadly. The response changed.
If this is automated, what do I do?
He suggested reallocating effort to analysis and higher-value work. The concern remained.
He raised the issue with his manager. The response was neutral.
You can share with the team.
You can work with the team.
There was no endorsement.
No role redesign.
No progression path.
No signal that improvement would be protected.
Improvement existed, but had no place to land.
Automation stopped where it threatened existing roles.
The new joiner disengaged. He had increased productivity and improved output.
The organisation had no structure to absorb it.
He requested a transfer to another department.
This was not resistance to change.
It was not a lack of initiative.
Productivity gains were permitted only where roles remained intact.
When improvement reduces work without redefining responsibility, it creates risk for those inside the system.
The organisation preserved roles. And removed the opportunity to improve.
This failure mode is one surface expression of a deeper structural loop.
The full model is developed in When Failure Becomes Rational.
When “data-driven” is blocked by design
Intent was declared. Execution never followed
The organisation announced a shift toward data-driven decision-making. A formal data function was established.
Two years later, a basic daily sales dashboard still did not exist.
The data function had no access to sales target data.
Sales targets sat outside the data warehouse.
The function had no authority to access source systems.
Authority to add the data sat with a senior leader who did not report to the data function.
The data function raised the dependency upward. It was rejected by middle management.
No escalation followed.
To keep delivery moving, the data function proposed a workaround.
Users would extract sales targets manually.
Reformat them.
Upload them weekly.
Check them repeatedly so the dashboard could run.
A user attempted to escalate the issue independently. Management did not respond.
Silence replaced decision.
The transformation team became involved.
The dependency was visible.
The constraint was understood.
Escalation to the authority that could redesign access did not occur. Instead, the same workaround was reinforced.
Manual extraction was added on top of the user’s existing responsibilities.
The process was documented.
Standardised.
Maintained.
Manual work became infrastructure.
Data responsibility was assigned.
Data authority was not.
The data function was accountable for insights.
It had no control over access.
No ability to compel integration.
No forcing mechanism when dependencies blocked delivery.
When escalation terminates below authority, redesign does not occur.
Workarounds do.
The organisation wanted data-backed decisions. The design made manual extraction permanent.
This failure mode is one surface expression of a deeper structural loop.
The full model is developed in When Failure Becomes Rational.
When incentives reward visibility over impact
Manual work and data issues emerged after automation was declared complete.
Company HQ rolled out a tool intended to automate an existing manual process.
On paper, the process was common across markets.
In reality, distribution structures, complexity, and required levels of detail varied.
Adopting the tool required more than training.
It required changes to ways of working, system adjustments, and coordination across multiple functions.
Two transformation leads took different approaches. The first prioritised speed and visibility.
Demonstrations were conducted.
Adoption was mandated.
Markets were instructed to use the tool as delivered.
To comply, markets force-fit their existing processes into the system.
Manual steps increased.
Workarounds multiplied.
Additional reconciliation was required to correct mismatches created by the forced fit.
Productivity declined.
Data quality issues surfaced.
Despite this, adoption metrics were met — quickly.
The second lead moved more slowly.
He engaged markets to understand how work actually flowed.
He coordinated alignment across functions.
He pushed through changes to ways of working and supporting systems.
That coordination depended on influence, enabled by direct access to local market CFOs and General Managers.
Both approaches produced visible adoption. Only one reduced manual work.
By the time the consequences of the first approach became visible, the lead responsible had already been promoted.
The costs — increased workload, degraded data, and remediation effort — were borne by local markets and the remaining transformation teams.
There were no repercussions for the decision path that created the damage.
This was not a failure of effort.
And it was not a failure of intent.
It was incentive misalignment.
When speed and visibility are rewarded without accountability for downstream consequences, systems learn to optimise for optics rather than deliver real work.
This failure mode is one surface expression of a deeper structural loop.
The full model is developed in When Failure Becomes Rational.
When strategy fragments at execution
Partial owners set to compete where collaboration was required.
A Vice President identified an under-targeted customer segment as a growth priority.
Execution responsibility was assigned to three team heads from different functions. Each team owned only a partial view of the customer problem.
No owner was appointed to integrate those views.
All three teams depended on the same research and analytics group for insight generation and execution support.
The group consisted of five people — three permanent staff and two contractors — responsible for business-as-usual reporting, ongoing analytical support, and multiple other initiatives and projects across the organisation.
Capacity was fixed.
Demand was triplicated.
Each team pursued the strategy independently, requesting analysis through its own functional lens.
The research team executed all three streams in parallel, while continuing to service unrelated work.
Work duplicated. Priorities conflicted. Trade-offs went unmanaged.
None of the three initiatives succeeded.
This was not miscommunication.
And it was not a capability issue.
Failure followed from design.
An integrated problem was split across partial owners and set in competition.
None had complete knowledge of the customer segment.
When intent is singular but ownership is fragmented across incomplete views, execution cannot converge.
It fragments by design.
This failure mode is one surface expression of a deeper structural loop.
The full model is developed in When Failure Becomes Rational.
When execution is blocked by the wrong decision rights
Authority was assigned. Accountability did not follow.
A regulator mandated a change to industry classification that affected a monthly regulatory report.
Data capture processes were redesigned. Source systems were updated. The data warehouse was adjusted. Downstream reports were rebuilt. End-to-end testing was completed.
A large, cross-functional effort.
When the work was finished, only formal sign-off remained. The regulatory deadline was days away.
Decision authority sat with a mid-level manager who had no involvement in the work and limited domain expertise across the changes made.
He did not review the work. He did not decide. He attempted to pass the decision elsewhere.
Responsibility circulated. No one decided.
Delay carried no cost. Action carried risk. Non-decision became the safest option.
Execution resumed only after the formal decision path was bypassed. Additional analysis was prepared. Another senior stakeholder was engaged and convinced to assume responsibility.
The change was approved. The deadline was met. But the design held.
Authority sat with someone detached from the work and insulated from consequences.
He held veto power without domain knowledge. He absorbed no risk if the deadline was missed. He carried full exposure if the change was wrong.
When sign-off power is detached from knowledge and accountability, urgency does not accelerate execution.
Formal authority collapses, and decisions occur outside the system.
This failure mode is one surface expression of a deeper structural loop.
The full model is developed in When Failure Becomes Rational.