Why the saving that gets negotiated and the cost that gets created are rarely in the same spreadsheet
There is a conversation that happens in most manufacturing businesses, in some form, every year.
Procurement presents its annual savings report. The numbers are good. Unit costs are down. Supplier consolidation has delivered scale. Negotiation has produced better terms. The total saving is significant — millions of pounds, depending on the size of the business. The function has delivered against its targets.
Operations has a different experience of the same year.
A supplier who reduced their unit price also reduced their quality investment. Incoming inspection failures increased. A production line that ran at 96% schedule adherence twelve months ago is now running at 88%. Premium freight to cover short deliveries has absorbed a quarter of the procurement saving. A quality escape that made it to a customer generated a warranty claim that absorbed another quarter.
The procurement report and the operations reality describe the same year.
They are not recognising the same costs.
This post is about the structural gap between how procurement decisions are made and the operational cost they create — and about what a genuinely integrated approach to supply chain economics looks like.
1. The Unit Cost Trap
The dominant metric in most procurement functions is unit cost.
Price per component. Cost per kilogram. Rate per service. The negotiated reduction from the previous period’s rate.
Unit cost is a real metric. It is relevant. It is not sufficient.
The unit cost trap is the logical error of optimising for the purchase price of a component without accounting for the total cost of the component’s presence in the value chain.
The total cost of a purchased component includes: the purchase price, the cost of receiving and inspecting it, the cost of any quality failures it generates in production, the cost of any supply failures it generates in scheduling, the inventory that must be held to buffer against its unreliability, the premium freight required when it arrives late or short, the warranty cost if quality failures reach the customer, and the management time spent maintaining a supplier relationship that is not performing.
In most supply chains, the purchase price is the largest component of that total cost. But it is not the only one. And for the components and suppliers where supply risk and quality risk are high, the non-price components can be substantial — sometimes exceeding the purchase price itself.
Optimising only for purchase price, when the total cost includes all those other elements, produces decisions that look right on the procurement scorecard and feel wrong in the factory.
2. How the Adversarial Model Forms
The adversarial relationship between procurement and operations in most manufacturing businesses is not the result of bad people making bad decisions.
It is the result of a measurement architecture that puts the two functions in structural conflict.
Procurement is measured on cost reduction. Cost reduction is most directly delivered by reducing unit price. Unit price is most directly reduced by competitive tendering, by supplier consolidation to drive scale, and by negotiating harder in annual price reviews.
Operations is measured on schedule adherence, quality, and cost of production. All three of those metrics are affected by supplier performance. When a supplier delivers late, operations misses schedule. When a supplier delivers defective components, operations absorbs scrap and rework. When a supplier’s reliability forces operations to carry higher inventory or pay premium freight, operations absorbs the cost.
The saving sits in the procurement metric. The cost sits in the operations metric.
Nobody has designed a single metric that captures the net effect.
So procurement optimises for its measure. Operations absorbs the consequence. And the tension between them — expressed in escalation meetings, in supplier review arguments, in the monthly operational review where premium freight and schedule adherence and quality escapes are explained — is a structural feature of the measurement system, not a failure of individual behaviour.
3. The Supplier Relationship as Operational Infrastructure
There is a model of the supplier relationship that is fundamentally different from the one most businesses operate.
In the transactional model — which is what most businesses have, regardless of what their procurement strategy documents say — the supplier is a vendor. They provide a product or service at a specified price. They are held to a specification. When they fail against the specification, they are penalised. When their price can be reduced by switching or threatening to switch, it is.
In the partnership model — genuinely, not rhetorically — the supplier is part of the value chain. Their process capability is the business’s process capability. Their quality system is the first line of quality defence. Their delivery reliability is a direct input to production schedule adherence. Their investment in process improvement is an investment in the final product’s cost and quality.
Toyota’s approach to supplier relationships — developed over decades and extensively studied — is the closest real-world example of the partnership model at scale. The key features are: long-term commitment that gives suppliers the confidence to invest, deep technical engagement that develops supplier capability, transparent cost visibility rather than adversarial negotiation, and mutual improvement objectives that align supplier investment with buyer outcome.
This is not a soft model. It is a rigorous one. It requires the buyer to understand the supplier’s cost structure in genuine detail. It requires the supplier to open their operations to scrutiny. It requires both parties to invest in the relationship over a time horizon that extends beyond the annual price review.
Most businesses have neither the patience nor the supplier development capability to operate at this level with their full supply base. But most businesses have a small number of critical suppliers — the ones whose performance most directly affects operational outcomes — for whom the partnership model would generate significantly better returns than the transactional one.
The question is whether the commercial and operational will exists to make that transition.
4. Total Cost of Ownership as Design Principle
The concept of Total Cost of Ownership — TCO — is not new. It has been discussed in procurement literature for decades.
It is rarely applied rigorously in practice, for a simple reason: calculating the full TCO of a supply relationship requires data that sits across multiple functions and systems, and nobody owns the integration of that data.
Procurement owns the purchase price data. Operations owns the schedule adherence and quality data. Finance owns the premium freight and inventory carrying cost data. Nobody owns the sum.
Building a genuine TCO model for key supply relationships requires a cross-functional data integration that most businesses haven’t invested in — and a shared accountability for the total that cuts across the functional measurement architecture.
But the return on that investment is significant.
When a business can see the full cost of a supply relationship — not just the unit price but the quality cost, the schedule cost, the inventory cost, the management cost — procurement decisions look different. The cheapest supplier is often not the lowest-cost supplier once the full picture is visible. The supplier who charges five percent more but delivers with 99.5% OTIF and near-zero defect rates is generating a different total cost than the one who charges five percent less and delivers at 87% OTIF with a two percent defect rate.
TCO visibility doesn’t mean the decision always favours the more reliable supplier. Sometimes the risk profile and the price differential mean the cheaper option is genuinely right. But TCO visibility means the decision is made with full information rather than partial information — and that decisions made with full information are, on average, better.
5. The Make-Versus-Buy Illusion
Every manufacturing business makes make-versus-buy decisions.
Some things are made in-house. Others are purchased. The boundary between them is set by a combination of strategic intent, historic capability investment, and economic calculation.
The economic calculation is almost always based on unit cost comparison: does it cost less to make this ourselves or to buy it from a supplier?
That calculation is systematically incomplete, for the same reason the procurement unit cost metric is incomplete.
The cost of making in-house includes: material, direct labour, machine time, tooling, overhead allocation.
The cost of buying includes: unit price — and sometimes the cost of managing the supply relationship, incoming inspection, supplier development, and the operational risk associated with external supply.
The cost of buying almost never includes the operational consequences of supply variability. The schedule disruption when the supplier misses a delivery. The quality cost when the supplier ships a non-conforming batch. The inventory investment required to buffer against supplier unreliability.
When those factors are included in the make-versus-buy calculation, the case for insourcing a strategic component is often stronger than the unit cost comparison suggests. And the case for outsourcing a non-strategic one is often stronger than the operational team’s preference for control implies.
The make-versus-buy decision made with full cost visibility is frequently different from the one made on unit cost alone.
6. Procurement as Operational Function
The structural response to everything described in this post is a design change, not a process change.
Procurement needs to be redesigned as an operational function — one whose primary measure of success is the operational performance of the supply base, not the unit cost reduction it generates.
That means: measuring procurement on supplier OTIF, not just on price. On incoming quality, not just on specification compliance. On the total cost of supply relationships, not just the purchase cost. On the operational capability of key suppliers, not just their commercial terms.
It means: involving procurement in operational improvement programmes — in supplier gemba walks, in value stream mapping that extends upstream into the supply chain, in the kaizen events that are addressing quality and schedule performance issues whose root cause sits in a supplier’s process.
It means: giving operations a formal input into supplier selection and supplier performance management — not just an escalation route when things go wrong, but a structured voice in decisions that will determine their operational environment.
And it means: building the capability in the procurement function to think operationally. To understand process capability and what it means for component quality. To read a supplier’s delivery performance data and identify the root causes of variability. To conduct a supplier audit not as a compliance exercise but as a genuine operational assessment.
That is a different capability from commercial negotiation and contract management. It requires either developing it in existing procurement teams or building a supply chain function that integrates commercial and operational expertise in a way that neither procurement nor operations typically does independently.
7. The Lean Supply Chain
Lean thinking applied to the supply chain is not about making suppliers Lean.
It is about making the extended value chain — from raw material to customer delivery — flow with minimum waste and maximum reliability.
That requires: demand signals that are accurate and stable enough that suppliers can plan without excessive buffer. Replenishment rhythms that are predictable and pull-based rather than batch-driven and variable. Quality assurance that is built into the supplier process rather than inspected at the receiving dock. Logistics that is designed around flow rather than around minimising transport cost in isolation from its operational consequence.
Each of these requires collaboration that the transactional procurement model doesn’t support. The supplier needs to know the customer’s real demand signal, not a purchasing order schedule designed to keep the buyer in control. The logistics model needs to be designed jointly, not specified by one party and quoted by another.
This is the direction that the most advanced manufacturing businesses are moving. Not because it is ideologically appealing, but because it produces measurably better operational outcomes — lower total cost, higher service reliability, faster response to market change.
It requires a different kind of procurement leadership. One that sees its role as supply chain architect rather than commercial negotiator.
Final Thought
The saving that procurement negotiates and the cost that procurement creates are rarely in the same spreadsheet.
That is a design failure, not a people failure. The measurement system creates the separation. The organisation structure reinforces it. The management process never integrates the two views into a single picture.
Closing that gap — building the TCO visibility, redesigning the measurement architecture, repositioning procurement as an operational function — is not a procurement improvement programme.
It is a leadership decision about how the business is designed to make supply chain decisions. About whether the operational cost of commercial choices is visible to the people making those choices. About whether the supply base is managed as a strategic asset or as a cost reduction opportunity.
The businesses that have made that decision consistently outperform those that haven’t.
Not because their suppliers are better.
Because they manage them more intelligently.
Three questions.
What is the full total cost of your five highest-spend supplier relationships — including quality, schedule, inventory, and premium freight costs, not just unit price?
When did procurement last participate in an operational improvement event whose root cause sat in the supply base?
And if you redesigned your procurement function’s success metrics tomorrow — what would you measure, and how different would the resulting behaviour be?
Leave a Reply