The waste your P&L doesn’t label as waste
No, this is not an April Fools’, but this can make a fool out of a business leader if you are not aware!
Every manufacturing business has two factories.
The first one is on the floor plan.
It has machines, cells, assembly lines, test rigs. It has cycle times and OEE and changeover schedules. It has operators and team leaders and shift managers. It is measured obsessively. It is the subject of improvement programmes, capital investment decisions, and Gemba walks. It is where Lean lives, officially.
The second factory has no floor plan.
It lives in the overhead base. In the functions. In the management layers. In the governance processes and the approval chains and the meeting calendars and the reporting cycles and the systems that talk to other systems imperfectly and require people to reconcile the difference.
It produces nothing of value.
It consumes an extraordinary amount of resource.
And almost nobody is looking at it through a Lean lens.
This post is about that second factory. About how it forms, why it grows, what it costs, and what it actually takes to redesign it — not with a cost-cutting programme, but with the same rigour you’d apply to eliminating waste from a production line.
1. What Overhead Actually Is
Start with the definition, because it matters.
Overhead is often described as fixed cost. The implication is that it is stable, structural, and largely unavoidable — the price of running a business of this complexity at this scale.
That framing is wrong. And it is expensive.
Overhead is not fixed. It is accumulated.
It is the sum of every decision — about structure, process, governance, system, and headcount — that was made over years and decades, often for good reasons at the time, and rarely revisited when those reasons changed.
The project management office that was created to manage a specific integration and never formally disbanded. The reporting pack that was designed for a previous MD’s preferences and is still produced monthly for an audience that no longer exists. The approval process that required three sign-offs because of a compliance failure five years ago that has since been addressed through other means. The regional structure that made sense when the business operated in four distinct markets and now mostly coordinates businesses that have already integrated.
None of those things are malicious. None of them were designed to waste money. Each one was a rational response to a specific circumstance.
Together, they constitute the hidden factory.
2. How It Grows
The hidden factory doesn’t appear overnight.
It accumulates through entirely predictable mechanisms, and understanding them is the first step to doing something about them.
Complexity addition without complexity removal.
Every new product, customer, market, or regulatory requirement adds process complexity. New approvals, new documentation, new reporting, new coordination. The business becomes more capable of handling more variety. But the processes that were designed for the old variety rarely get retired. They coexist. The system grows laterally — never backward.
A business that handles twice the product variants it did a decade ago will typically have more than twice the overhead, because it added the new overhead to manage new complexity without removing the old overhead that new complexity made redundant.
The safety layer phenomenon.
When something goes wrong — a quality escape, a financial misstatement, a customer complaint, a regulatory finding — the organisational response is almost always additive. A new check. A new approval gate. A new review. A new sign-off.
These responses are rational. The check prevents the problem recurring. The approval gate adds scrutiny. The review provides oversight.
But they are cumulative. And they are almost never removed when the underlying problem has been structurally resolved.
Over time, the business develops a thick layer of governance that was built incrementally in response to specific historic failures and that now applies universally, whether the risk justifies it or not.
Functional empire building.
Functions grow. It is in their nature.
Not because of bad intent — because of good intent. Every functional leader believes their team is understaffed relative to the demands placed on it. Every business case for headcount is made with genuine justification. Every new capability is genuinely needed at the point it is added.
But the aggregate — the total headcount across all functions, the total cost of coordination between them, the total overhead generated by their interaction — is rarely assessed as a system.
Each function is optimised locally. The whole is never designed.
System proliferation.
ERP. MES. CRM. WMS. PLM. APS. QMS.
Most businesses of any scale are running eight to fifteen distinct systems. Each was implemented to solve a specific problem. Each solved it, partially. Each requires data that exists in another system and doesn’t flow automatically. Each therefore requires people — analysts, planners, coordinators — to extract, reconcile, translate, and reformat data between systems.
The hidden factory runs largely on spreadsheets. Not because people are unsophisticated, but because the systems that were supposed to eliminate manual work created new categories of it instead.
3. What It Costs
The honest answer is that most businesses don’t know.
Because the hidden factory’s costs are distributed across functional budgets, absorbed into salary lines, embedded in system licences, and treated as structural rather than operational. Nobody has ever added up what it actually costs to run the second factory.
But the components are real and they are measurable.
The coordination tax.
Every hour spent in a meeting that exists to reconcile information that should flow automatically. Every email chain that is resolving an ambiguity that a well-designed process would have eliminated. Every escalation that is happening because accountability is unclear. Every cross-functional conversation that is necessary because two functions are making decisions in parallel that affect each other.
In most mid-sized manufacturing businesses, senior operational leaders spend between thirty and fifty percent of their time on coordination activity. Coordinating between sites, between functions, between systems, between time zones, between planning cycles.
That is thirty to fifty percent of the most expensive cognitive resource in the business, consumed by the friction of a poorly designed system rather than applied to the problems the system was designed to solve.
The rework loop.
Every report that is produced and then corrected. Every plan that is built in one system and then manually adjusted in another. Every data reconciliation that exists because two systems don’t share a common definition of the same term. Every governance process that requires a formal submission, a review, a revision, and a resubmission.
White collar rework is the most invisible waste in manufacturing. It doesn’t generate scrap. It doesn’t trigger a quality alert. It appears nowhere in the OEE calculation. But it consumes time, creates latency, and generates the kind of low-level frustration that erodes both engagement and decision quality.
The latency cost.
When information has to pass through multiple layers before it reaches the person who can act on it, decisions are slow. When decisions are slow, the window for good options closes and the range of available responses narrows.
The business that can make a supply decision in twenty-four hours has a structural advantage over one that takes four days. Not because it has better information — often the information is identical — but because its overhead architecture routes information efficiently rather than processing it through unnecessary intermediaries.
Latency is a cost. It is a competitive cost. It is directly produced by the hidden factory.
The talent cost.
The hidden factory consumes capable people doing work that adds no value to the customer.
That is not a trivial observation. Those people were hired, trained, and retained at significant cost. They are engaged in coordination, reconciliation, reporting, and governance that exists because the system requires it. They are not stupid and they are not lazy. They are trapped in processes that could be redesigned.
The organisations that have eliminated significant hidden factory cost haven’t typically made people redundant. They have redeployed them. The analyst who spent sixty percent of her week reconciling data between two systems now has sixty percent of her week available for actual analysis. The improvement that delivers that outcome isn’t headcount reduction. It is capability liberation.
4. Why Nobody Talks About It
If the hidden factory is this costly and this visible once you know where to look, why is it so rarely the subject of serious leadership attention?
Three reasons.
It’s not on the Gemba walk.
Lean methodology is overwhelmingly oriented toward the shop floor. The tools — value stream mapping, 5S, standard work, OEE, SMED — were designed for physical production processes. They translate imperfectly to functional work. A VSM of the month-end close process or the engineering change workflow is possible, but it requires a different kind of facilitation and a different kind of courage.
Leaders who are excellent operational managers often have a blind spot for functional waste because their lens was developed on the factory floor and hasn’t been extended to the office.
The waste is socially embedded.
Eliminating a production step that adds no value is operationally straightforward. Eliminating a governance process that involves twelve people, three functions, and a management layer is politically complicated.
Those twelve people have roles that are partially or wholly constituted by that process. Their managers have accountability structures built around it. The process has history — it was created for a reason, possibly by someone still in the building. Challenging it is challenging something someone built, which is challenging someone.
The hidden factory persists partly because redesigning it requires a directness that most organisations are not culturally equipped for.
It gets attacked with the wrong tool.
When overhead becomes a problem — usually because margin is under pressure and the board wants the cost base reduced — the response is almost always a cost-cutting programme. Headcount targets. Budget compression. Discretionary spend freezes. Supplier negotiations.
These work, briefly. And then they stop working.
Because they remove resource from the system without removing complexity. The people who did the coordination are gone, but the need for coordination remains. The reports that were eliminated come back in a different form six months later because someone still needs the information. The approval gate that was abolished gets reinstated after the first incident that it would have caught.
Cost-cutting treats overhead as volume. Lean treats it as design.
The hidden factory cannot be defunded out of existence. It has to be redesigned out of existence.
5. Seeing the Hidden Factory
Before you can redesign it, you have to see it.
That requires a different kind of observation than the Gemba walk — but the same underlying discipline. Go to where the work actually happens. Watch what people actually do. Ask why each step exists and what would happen if it didn’t.
The signals that the hidden factory is large and costly are not subtle once you know what to look for.
The meeting-to-decision ratio.
Count the meetings in your leadership calendar over the past month. Then count how many of them produced a material decision — one that changed something, committed resource, or resolved a question that had been open. In most organisations, that ratio is poor. Meetings exist to share information, align understanding, review performance, and prepare for other meetings. The decision density is low.
A high volume of meetings with low decision density is a diagnostic indicator of a system that hasn’t clarified where accountability sits. When everyone needs to be informed before anything can happen, the process has been designed around consensus rather than accountability. That design is expensive.
The spreadsheet census.
Ask your finance, planning, operations, and commercial functions how many active spreadsheets they maintain — files that are regularly updated and shared as part of live business processes. The number is almost always surprising. In most businesses it runs into the hundreds. Many of those spreadsheets exist because a system doesn’t do something it was supposed to do, or because two systems don’t talk to each other, or because someone needed information in a format that no system produces automatically.
Each spreadsheet is a process. It has an owner, a maintenance burden, an update cycle, and a downstream dependency. It is, in Lean terms, a workaround. It represents a system design failure that has been solved locally rather than structurally.
The approval chain analysis.
Take five recent decisions — a capital expenditure, a contract, a product change, a supplier qualification, a customer exception — and map the approval path each one actually took. Not the formal policy version. The actual path: who looked at it, who added comments, who forwarded it to whom, how many rounds of revision it went through, how long it took from initiation to decision.
Then ask: at each step, was genuine value added? Did the review change the decision, improve the quality of the analysis, or add accountability that wasn’t already present? Or did it add time, formatting work, and the organisational cost of keeping multiple people informed of something that only one of them needed to decide?
The answer is almost always uncomfortable.
The report graveyard.
Identify every regular report produced in your business — daily, weekly, monthly, quarterly. Then find out who actually reads each one, what decisions it informs, and what would happen if it stopped being produced tomorrow.
A significant proportion of regular reporting in most organisations is produced because it has always been produced. The original audience has changed. The decision it informed is now made differently. The system that couldn’t produce the data automatically has since been replaced. But the report continues, because cancelling it requires someone to notice that it’s redundant and decide to stop, and neither of those things happens naturally.
6. The Lean Framework for Overhead
Apply the same logic to functional work that you apply to operational work.
Define value from the customer’s perspective. For an overhead process, the customer is the person or function that uses the output. The question is: what does this output enable them to do that creates value for the end customer? If the answer requires more than one logical step, the process may be further from value than it appears.
Map the flow. A value stream map of a functional process is illuminating in exactly the way a production VSM is — it makes visible the wait times, the rework loops, the batching, the handoffs, and the steps that add no value. The ratio of value-added time to total elapsed time in most functional processes is as poor as the worst production processes in the building. Often worse.
Identify and eliminate the eight wastes. They all exist in overhead work.
Overproduction: reports produced that nobody uses. Information prepared at a level of detail nobody needs.
Waiting: approvals queued behind a busy inbox. Decisions deferred because the right person is unavailable.
Transport: information passed through unnecessary intermediaries before reaching the person who acts on it.
Over-processing: governance applied at a level of rigour that the risk doesn’t justify. Analysis produced to a precision that the decision doesn’t require.
Inventory: work queued in inboxes, in approval systems, in shared drives. Decisions pending. Reports prepared but not yet distributed.
Motion: searching for information across multiple systems. Reformatting data from one tool to another. Chasing responses through multiple channels.
Defects: errors in reports that require correction. Submissions that are rejected and resubmitted. Decisions that are reversed because they were made on incomplete information.
Underutilised talent: capable people spending the majority of their time on low-value coordination rather than high-value analysis, insight, or improvement.
Pull rather than push. Many functional processes run on a push logic — reports are produced and distributed on a schedule, regardless of whether the recipient needs the information at that moment. A pull-based functional process produces information when it is needed, at the level of detail that is needed, in the format that enables the decision. That is a fundamentally different design.
7. Where to Start
The hidden factory cannot be eliminated in a single programme. It took years to build and it will take sustained effort to redesign. But it can be addressed systematically, with a clear prioritisation logic.
Start with the processes that sit closest to value-creating decisions.
The planning processes that determine what gets made and when. The approval processes that govern capital and resource allocation. The review processes that determine where leadership attention goes. These are the overhead processes with the highest leverage — the ones where latency, rework, and poor design most directly affect the outcomes the business cares about.
Apply a simple test to each one.
What decision does this process enable? Who makes that decision? What information do they actually need? How close is the current process to providing exactly that and nothing else?
The gap between the current process and the answer to those questions is the waste.
Then apply the standard Lean improvement logic. Redesign the process around the value it is supposed to deliver. Remove every step that doesn’t contribute to that value. Clarify the accountability that currently sits ambiguously across multiple functions. Build in feedback mechanisms so that process performance can be measured and improved over time.
This is not glamorous work. It does not generate the visible drama of a kaizen event on the shop floor. But it compounds.
A functional process that takes four days and requires seven people can often be redesigned to take one day and require three. That improvement doesn’t show up as a quality metric or an OEE gain. It shows up as faster decisions, better analysis, and more capable people doing more valuable work.
Over three years, that compounds into a materially different overhead cost structure — not because costs were cut, but because the work was redesigned.
8. The Structural Question
There is a harder conversation sitting behind the process question.
Organisational structure is itself an overhead design choice.
The number of management layers. The span of control at each level. The degree of functional centralisation versus site autonomy. The decision rights that sit at group versus division versus site. The extent to which shared services have actually created sharing rather than just centralised duplication.
These structural choices determine the baseline cost of coordination in the business. They determine how many people are required to maintain alignment across the organisation. They determine how fast information moves and how many hands it passes through.
Most organisational structures were not designed. They evolved. Acquisitions added entities that were never fully integrated. Reorganisations addressed specific problems without considering the coordination costs they created elsewhere. Individual leaders built team structures that reflected their personal working style rather than the requirements of the work.
The result is an organisational architecture that is partly designed and partly accidental, and that carries significant hidden cost in the form of coordination overhead that exists because the structure requires it.
Redesigning structure is the highest-leverage intervention in the hidden factory. It is also the most difficult — politically, emotionally, and practically. It requires a clarity about accountability that many organisations are not ready for and a willingness to make decisions that affect people’s roles and seniority in ways that operational process improvement does not.
But it is worth naming. Because the organisations that have genuinely transformed their overhead cost structure have almost always addressed structure as well as process. Process improvement alone hits a ceiling defined by the structure it operates within. Structural redesign removes that ceiling.
9. The Multi-Site Dimension
In a multi-site business, the hidden factory has additional layers.
Every site develops its own local overhead — its own reporting rhythms, its own governance conventions, its own functional processes, its own informal coordination mechanisms. These are rational local adaptations. They are also a source of duplication.
Three sites each maintaining their own planning function, their own finance reporting process, their own HR operations, their own quality management system — each doing similar work, in slightly different ways, with slightly different tools — represents overhead that is at least partially reducible without any loss of capability.
But the duplication is usually invisible at group level. The site cost base is reported as a total. The functional breakdown within it is rarely compared systematically across sites. The question “are we paying for this three times?” is rarely asked, because the answer requires a level of functional transparency across sites that doesn’t naturally exist in most divisional structures.
The hidden factory in a multi-site business is not just the waste within each site’s overhead. It is the structural duplication between them.
And it grows significantly during periods of acquisition and integration — when new entities are absorbed with their own systems, processes, and governance conventions, and when the pressure to maintain continuity means that rationalisation is deferred indefinitely.
10. The Cultural Prerequisite
Everything in this post requires one thing that cannot be assumed.
A leadership team that is willing to examine its own processes with the same rigorous curiosity it applies to the shop floor.
That is harder than it sounds.
The Gemba walk on the factory floor is an established ritual. Leaders are comfortable examining production processes, asking questions, challenging assumptions, and demanding improvement. That behaviour is culturally legitimate.
Applying the same lens to a management process — to the month-end review, to the capital approval process, to the structure of the executive leadership team — is not yet culturally legitimate in most organisations. It feels like a criticism of the people who designed those processes. It often is a criticism of the people who are still running them. And those people are frequently in the room.
The organisations that have successfully attacked the hidden factory have leaders at the top who are willing to start with their own processes. Who will say: our planning cycle takes twelve days and requires thirty people, and that is a design choice we made, and we are going to redesign it. Who will apply to their own team’s work the same question they ask of the production line: what is the value-added time as a percentage of the total time this takes?
That takes a particular kind of confidence. The confidence of a leader who is not defensive about imperfection, who treats waste in the management system as an improvement opportunity rather than a reflection on their leadership, and who understands that eliminating overhead waste is not a cost-cutting exercise — it is a capability investment.
When that posture exists at the top, the hidden factory becomes visible.
When it doesn’t, it stays hidden.
Comfortably, expensively hidden.
Final Thought
The Lean transformation that stops at the factory gate is only half a transformation.
The discipline that eliminates waste from production processes, that maps value streams and reduces changeover times and drives OEE toward world class — that same discipline applies with equal force to the processes that govern, coordinate, plan, and report across the business.
The hidden factory is not a metaphor.
It is a real set of activities, consuming real resource, producing no customer value, and sitting largely unexamined in the overhead base of almost every manufacturing business in the world.
The tools to see it are the same tools you already have.
The logic to redesign it is the same logic you already apply.
What is usually missing is not capability. It is the decision to look.
Three questions to sit with.
When did you last value stream map a management process rather than a production process?
What percentage of your senior leaders’ time is spent on coordination that exists because the system requires it, rather than on work that directly creates value?
And if you applied your standard waste elimination logic to your overhead base — the same rigour, the same curiosity, the same intolerance for steps that add no value — what would you find?
The answer is already in your building.
You just haven’t walked that Gemba yet.
adam
Leave a comment