The Age of Diluted Accountability in Supply Chains
Ask a simple question in most large Australian supply chain functions today. Who owns this? Watch what happens.
You will get a list. You will get an org chart. You will get a RACI that somehow makes three different people accountable for the same outcome. What you will not get is a name.
This is not an accident. It is the accumulated result of several trends landing at the same time: AI absorbing the analytical middle of every decision, the rapid rise of fractional and advisory roles, the duplication of systems and data sets across functions, and a culture of permanent pilots that never quite arrive at a conclusion. The net effect is what we are calling the age of diluted accountability.
It is showing up in every sector we work in. Retail, FMCG, hospitality, government, defence, infrastructure, health, and aged care. And most worryingly, it is showing up in the compliance-driven corners of organisations where accountability is supposed to be clearest of all.
This is an argument for doing something about it.
What diluted accountability actually means
Diluted accountability is not the absence of process. Most organisations have more process than they did five years ago, not less. It is the condition where ownership of an outcome is so distributed across roles, systems, advisors, and algorithms that nobody is genuinely answerable when things go wrong, and nobody has the authority to act decisively when things go right.
You see it in the way decisions get described. "We've aligned with stakeholders." "The model recommended this." "The policy says." "Procurement is working with the business." These are not decisions. They are descriptions of decision-making processes in which no individual has put their name on the outcome.
The shorthand test is simple. When a result is bad, can you point to the person who owns fixing it? When a result is good, can you point to the person whose judgement produced it? If the answer to either question is a committee, a workstream, a system, or an advisor, accountability is diluted.
AI has absorbed the analytical middle. Nobody replaced what it moved
The biggest structural change in supply chain decision-making over the past three years is the quiet industrialisation of AI across the analytical layer. Demand forecasting, inventory optimisation, supplier risk scoring, contract review, freight optimisation, store replenishment, workforce rostering. Tasks that used to sit with analysts and planners are increasingly performed by tools, often with no human in the loop until the decision is ready to be made.
This is not the problem. In our own AI philosophy at Trace, we have argued consistently that AI should own the analytical middle of most engagements. It is faster, more consistent, and better at handling the combinatorial complexity of modern supply chains than any human team. The value is real, and the direction of travel is right.
The problem is what happens at either end.
At the front end, someone still has to frame the question. What are we solving for? What constraints actually matter? Whose interests are we weighing? These are judgement calls that require commercial, operational, and ethical context. Too often, they have become a checkbox in a software configuration rather than a considered act of leadership. The model inherits the framing, and the framing inherits the blind spots of whoever set up the tool last.
At the back end, someone still has to make the call. The model produces a recommendation. A human is meant to review it, apply judgement, and act. But when an analyst receives fifty AI-generated recommendations a week, review collapses into rubber-stamping. When a senior leader receives a summary of the summary of the summary, the accountability for the outcome has been laundered through so many layers that nobody can honestly claim to have made the decision.
This is the accountability grey zone of AI-assisted decision-making. The algorithm did not decide. The analyst did not decide. The manager who approved the batch did not decide. The system that executed the order did not decide. And yet the decision was made.
When it goes wrong, the conversation that follows is revealing. Nobody can explain why the decision was made. Nobody is willing to say they owned it. The post-mortem produces better guardrails for the next time, but no named accountability for this time. The cycle repeats.
Fractional roles and the erosion of end-to-end ownership
The second trend compounding the problem is the rise of fractional, interim, and advisory roles across the senior layers of Australian organisations.
Fractional CFOs. Fractional heads of supply chain. Interim procurement leaders. Part-time chief data officers. Strategic advisors who sit somewhere between the board and the executive. Consulting partners who are functionally embedded for six months at a time.
Many of these arrangements are rational. Not every organisation can justify a full-time senior executive in every discipline. Fractional talent brings experience that smaller and mid-market businesses could never afford on a permanent basis. Interim executives stabilise functions during transition periods. Advisors inject outside perspective that insiders cannot provide.
The issue is not any one of these roles. The issue is the cumulative effect.
In a growing number of organisations, the senior supply chain function is now a patchwork of fractional and advisory contributors. The fractional head of supply chain is two days a week. The data strategy advisor is there for the transformation programme. The procurement consulting partner runs the tender process. The operations interim fills the gap until a permanent hire is made. The AI lead is a vendor-supplied specialist on secondment. Each person is capable. Each person is well-intentioned.
But nobody is carrying the full weight of the function across a multi-year horizon. Nobody is present for the full arc of a decision, from framing through execution through consequence. Nobody accumulates the scar tissue that real accountability produces.
Fractional leaders arrive with their own frameworks, make recommendations, and leave before the consequences land. Successors inherit a set of decisions they did not make and often do not fully understand. Knowledge walks out the door on a quarterly cycle. What remains is a function that looks well-resourced on paper and is quietly leaking accountability in practice.
This is not a criticism of fractional talent. Many of the best people we work with operate in fractional capacities, often because it is the only way to secure them. It is a criticism of the governance gap that surrounds them. Without a named permanent owner in each functional area, fractional contributions compound rather than resolve the accountability problem.
Duplicated systems and data sets: three versions of the truth
If you want to see diluted accountability in its most tangible form, look at the data landscape of any large Australian organisation.
Walk through the supply chain function of a typical mid-to-large enterprise and you will find an ERP that holds the master inventory record, a warehouse management system with its own view of stock, a bolt-on planning tool that pulls from both but reconciles to neither, a transport management system with yet another data structure, a demand forecasting tool that runs off a subset of the planning data, a supplier risk platform with its own view of supplier status, a sustainability reporting tool drawing from procurement extracts that are themselves drawn from the ERP, a data lake, a data warehouse, and a handful of department-level SharePoint libraries holding what people quietly refer to as the real numbers.
Every one of these systems was purchased to solve a problem. Every data set was created to answer a question nobody could answer before. Individually, these were rational investments. In aggregate, the result is a decision-making environment in which different parts of the organisation are literally looking at different realities.
When nobody trusts the same numbers, nobody can be held accountable for them. An executive who is asked to own the inventory position can reasonably point out that three different systems disagree on what the inventory position is. The accountability question becomes a data question, which becomes an integration question, which becomes a multi-year transformation programme. The original question of who owns the outcome is lost in the scaffolding.
The problem is made worse by shadow systems. When the official systems are not trusted, teams build their own. The planner's Excel model. The procurement team's Power BI dashboard. The category manager's working file. These shadow systems are often more accurate than the official ones, which is precisely why they exist. But they are invisible to governance, immune to audit, and dependent on the continued presence of the individual who built them. When that person leaves, the real view of the function leaves with them.
Good data architecture is not a technology problem. It is an accountability problem. A single source of truth for each operational metric forces the organisation to decide who owns that metric. Without it, ownership is negotiable, and accountability becomes a matter of whose spreadsheet gets to the boardroom first.
Science projects that never end
The fourth driver of diluted accountability is the proliferation of what we call permanent science projects. Pilots, POCs, trials, and initiatives that were meant to run for a defined period and produce a decision, but instead become part of the organisational furniture.
Every organisation we work with has them. The AI forecasting pilot that was scoped for six weeks and is now in month eighteen. The digital twin initiative that has generated three steering committee papers and zero operational changes. The procurement analytics platform that has been live for two years and is still described as being in proof-of-concept. The traceability trial that keeps getting its scope revised rather than concluded.
Science projects are not the problem. Unmanaged science projects are. A pilot without a decision deadline is not a pilot. A POC without a kill criterion is not a proof of concept. A trial that outlives its sponsor, its business case, and its original technology stack is not a trial. It is a liability dressed up as innovation.
The accountability cost of permanent science projects is specific and severe. Operational decisions get deferred because the pilot might change the answer. Investment in core capability is starved because the science project is expected to render it obsolete. Teams develop expertise in running pilots rather than in running operations. And the original sponsors who could have made the call to scale or kill have moved on, taking the decision authority with them.
Good governance of science projects is not complicated. Every pilot should have a named executive sponsor with the authority to scale or kill. Every pilot should have a decision deadline, not just a review date. Every pilot should have a kill criterion agreed in advance. And every pilot should produce a written decision at the deadline, with a named person on the paper.
Organisations that do this end up with fewer pilots and more operational improvement. Organisations that do not end up with science projects as a permanent substitute for decisions.
Even compliance is not safe
The most unsettling part of this pattern is that it is now showing up in compliance-driven areas, where accountability is supposed to be the clearest.
Modern slavery. Product safety. Food safety. Biosecurity. Critical infrastructure security. Cyber supply chain. Quality assurance. These are areas where regulators, boards, and the public expect a single named owner who can answer questions, explain decisions, and bear consequences. And yet, in engagement after engagement, we find that accountability in these areas has been quietly diluted by the same forces described above.
Ask who owns the modern slavery statement in a large retailer. The answer is typically "the ESG team, working with procurement, with input from legal, and endorsed by the supplier risk committee." Press on who actually owns the statement, who would be called into a regulator's office if that statement turned out to be materially misleading, and the answer gets vaguer.
Ask who owns supplier security assessment in a critical infrastructure operator. "Cyber reviews the technical controls. Procurement manages the onboarding. The business owner signs off the contract. Risk aggregates the findings." Ask who is accountable for the decision to onboard a supplier with residual cyber risk, and the answer becomes a committee.
Ask who owns product traceability in a food manufacturer. "Quality owns the system. Supply chain owns the data. IT maintains the platform. Operations runs the daily checks." Ask who would personally be called to account if a contaminated product reached consumers, and the answer is often a legal construct rather than a person.
This is dangerous. Regulators and courts do not accept committee-based answers when something goes wrong. They look for individuals. When they cannot find one, they create one, usually by default and usually someone who did not expect to be holding the liability.
Boards are beginning to notice. Directors of Australian companies are increasingly asking their executives a version of the same question: who is the named, single point of accountability for this compliance outcome? Where the honest answer is "nobody", directors are insisting on change. This is a welcome trend, and organisations that get ahead of it will find themselves in a better position than those who wait for a regulatory event to force the issue.
What good looks like
Reversing diluted accountability does not mean reverting to rigid hierarchies or abandoning AI, fractional talent, or innovation pilots. All of those are here to stay and, when properly governed, all of them are net positive. What it does mean is applying a small set of disciplines consistently.
A named owner for every outcome, not just every process. Processes have owners already. Outcomes often do not. The distinction matters. An outcome owner is the person who is personally answerable for the result, regardless of how many systems, functions, and advisors contributed to it. For every material operational outcome (service level, inventory accuracy, procurement savings, compliance posture, cost-to-serve), a named executive should be on the record as the owner.
AI positioned as decision support, not decision maker. The framing matters. AI that is positioned as the decision maker creates accountability ambiguity by default. AI that is positioned as decision support, with a named human decision maker on every material output, preserves accountability by design. This is particularly important for compliance-adjacent decisions.
Science projects with explicit decision deadlines and kill criteria. Every pilot should have a sponsor, a deadline, a success threshold, and a kill criterion agreed in advance. No pilot should outlive the executive who sponsored it without an explicit decision to continue.
A single source of truth for each operational metric. This is a technology problem, a data governance problem, and an accountability problem, in that order. Organisations that invest in data architecture that enforces single sources of truth find that accountability follows the data, not the other way around.
Compliance accountability vested in a named executive. Compliance-driven outcomes should be owned by a named person at the executive level, not by a committee. The committee can advise. The executive is on the hook.
None of this is revolutionary. What is remarkable is how rare it has become.
How Trace Consultants can help
Trace works with Australian organisations across retail, FMCG, hospitality, government, defence, infrastructure, and health and aged care to restore accountability across supply chain, procurement, and operations functions. Our approach deliberately combines senior practitioner leadership with AI-enabled analytical depth, reflecting our view that AI should own the analytical middle while humans own problem framing, judgement, and implementation.
Operating model and organisational design. We help organisations redesign supply chain and procurement operating models so that accountability is named, clear, and supported by the right structure, roles, and decision rights. See Organisational Design for how we approach this.
Planning, operations, and decision governance. We work with planning and operations leaders to restore clean lines of accountability across S&OP, inventory, forecasting, and service level decisions, including the governance layer around AI-enabled planning tools. See Planning and Operations.
Procurement accountability and category ownership. We redesign procurement functions so that category managers own outcomes, not just processes, and so that compliance obligations (modern slavery, supplier risk, ESG) sit with named owners rather than committees. See Procurement.
Technology and data architecture. We help organisations rationalise duplicated systems and data sets, establish single sources of truth for operational metrics, and right-size their supply chain technology stack. See Technology.
Resilience and risk governance. We work with boards and executive teams to name ownership for supply chain risk, compliance, and resilience outcomes, and to put the governance layer in place that supports it. See Resilience and Risk Management.
Project and change management. We take pilots, POCs, and transformation programmes and bring them to a decision. When a programme is stuck, the fix is rarely more analysis. It is clear governance, named accountability, and a deadline. See Project and Change Management.
Where to begin
Restoring accountability is less about transformation than it is about clarity. Most of the work is in the first four weeks.
Start by mapping the ten most material operational and compliance outcomes in the function. For each, ask a single question. Who is the named, permanent, accountable owner? Where the answer is a committee, a fractional role, a system, or a vendor, that is where the work is.
Then look at the science project portfolio. List every pilot, POC, and trial currently running. For each, ask two questions. When is the decision deadline? What is the kill criterion? Anything without clear answers to both questions should be paused until it has them.
Finally, look at the data landscape. Pick the three or four metrics that matter most to the function (service level, inventory accuracy, supplier performance, procurement savings, a compliance indicator). For each, identify the authoritative system and the authoritative owner. Where there is disagreement, resolve it at the executive level and write it down.
This is not a multi-year programme. It is a set of decisions that can be made in weeks. The organisations that make them recover something that gets harder to recover the longer it is left. The organisations that do not will continue to accumulate the costs of diluted accountability, in service failures, compliance events, and the quiet erosion of decisiveness that makes operational functions great.
The age of accountability is worth reclaiming
AI is not going away. Fractional talent is not going away. Systems will continue to proliferate. Pilots will continue to get funded. None of that is the enemy. The enemy is the quiet slide into a world where nobody owns anything, where every decision is a collective product of systems and advisors, and where the answer to "who is accountable" is a list.
Australian organisations that notice this early and reverse it will build a durable advantage. They will move faster. They will be more trusted by their regulators, their boards, and their customers. They will attract the kind of senior talent that wants to own outcomes, not manage committees. And they will find that many of the problems they thought were technology problems or transformation problems were accountability problems all along.
Diluted accountability is a choice, not an inevitability. It is worth the effort to make a different one.

















.jpg)











