Nobody Owns the Seams: Why Capex Committees Approve Projects but Not Systems

Opening Insight

The quarterly financial review runs on a Tuesday. The CFO is walking through variance against plan.

Labor cost is 9 percent over budget.

Overtime is elevated, up 18 percent year over year.

Two major capex projects approved 12 months ago are showing payback stretch. The case packer on Line 2 is at 31-month payback against a modeled 22. The scheduling initiative is underperforming against its business case.

Three separate meetings are scheduled to address the variances. The labor cost review goes to HR and Operations. The capex stretch review goes to Finance and the project sponsor. The scheduling underperformance review goes to the continuous improvement team.

The CFO stops mid-walkthrough. Same root cause might be surfacing across all three. She asks: who do I call about this?

The room pauses. Ops owns utilization. Finance owns payback. HR owns headcount. Procurement owns sourcing. Each owner has their piece. No one owns the interaction.

The gap is not a missing model. It is a missing role.

This is the pattern across mid-market manufacturers. The capex committee has sixteen people in it. Not one of them is accountable for whether the sum of approved projects delivers the sum of projected savings. The committee approves components. No one approves the system.

System Context

Org charts are decision-scoped. Each of the Four Decisions has a functional owner.

At a typical mid-market manufacturer:

• Labor: VP of Human Resources, with operational partnership from the Plant Manager
• Automation: VP of Operations, often with CFO co-ownership on capex
• Scheduling: Director of Operations or Director of Continuous Improvement
• Packaging sourcing: VP of Supply Chain or Chief Procurement Officer

Each wins or loses on their own metric. The Mechanism below shows what happens at the seam.

Mechanism

A job description specifies the decisions an owner is accountable for. No job description specifies the interactions an owner is accountable for. The accountability gap is structural, not a failure of individual performance.

The mechanism is easier to see by walking through a single interaction.

When a packaging sourcing change lands and the line's throughput drops, who owns the diagnosis?

Procurement delivered the unit-cost savings. Their metric hit. They closed the file.

Operations sees the throughput drop. They open a line issue. Maintenance gets involved. Quality gets involved.

The throughput drop is ultimately traced back to the film spec change. But by this point the material is running, the contract is signed, and the procurement savings are in the Q2 numbers. Who owns the reversal decision? Procurement's job description says "deliver unit-cost savings," not "reverse completed sourcing changes when downstream throughput drops." Operations' job description says "run the plant," not "second-guess Procurement's wins."

The resolution happens through escalation. Someone senior enough to have authority over both functions has to decide. That is usually the CFO or COO, who is also the only person in the org who can see both the procurement savings line and the throughput loss line on the same page.

Between the identification of the problem and the CFO getting involved, typically four to six weeks pass. That time is the cost of the unowned seam.

System Interaction

The unowned seam shows up differently in each of the Four Decisions' interactions, but the pattern is consistent.

Labor and Scheduling: when the mix drifts and overtime rises, HR and Ops each see their piece. HR sees hiring pressure. Ops sees schedule adherence. Neither sees the mix-drift root cause. The fix, rebuilding the labor plan, requires both functions to pause their individual metrics and realign. There is no forcing function.

Automation and Scheduling: when a capex project underperforms, the project sponsor (usually in Ops or Finance) owns the review. The scheduling assumption embedded in the capex case is not anyone's active accountability. It lived in the appendix. The review stays at the equipment layer.

Sourcing and Labor: when a material spec change adds operator touches, Procurement's savings are counted and HR's staffing plan stays unchanged. The added labor cost absorbs in overtime, which gets flagged as a labor problem, not a sourcing consequence.

Symptoms get assigned to the owner of the symptom, not the owner of the cause. Because the cause lives at the seam, the cause has no owner.

Economic Consequence

Approved on clean math. Running on messy reality.

The cost of an unowned seam is not a single line on the P&L. It is a set of recurring overheads:

• Escalation overhead: an average of four to six weeks of cross-functional escalation per significant interaction issue, typically 4 to 8 events per year at a mid-market manufacturer. Aggregate leader time cost: 20 to 40 weeks annually
• Duplicate resolution: the same root cause gets addressed separately by three or four teams, each running its own meetings, each partially reinventing the analysis
• Recurring pattern: because the seam is unowned, the next year's capex cycle produces the same pattern: new projects, new gaps, new escalations
• Missing role: a VP of Operational Integration (or equivalent) typically costs 200 to 400 thousand fully loaded. The Silo Tax it would eliminate typically runs 3 to 5x that on a mid-market manufacturer

The cost of an unowned seam is compounding. Each year's capex cycle produces the same failure pattern because the org design that produced last year's gap is still in place. The Silo Tax is structurally embedded, not operationally accidental.

The P&L does not show this as a line. It shows the symptoms: elevated overtime, capex stretching, sourcing wins that did not materialize. The root cause does not have a home on the income statement.

Diagnostic

The diagnostic is organizational, not operational. One question:

Which job title on your org chart is accountable for whether the sum of approved capital projects delivers the sum of projected savings?

If the answer is a specific person with a specific role, the organization owns the seam.

If the answer is "the committee" or "we all do" or "no one specifically," the seam is unowned.

A second test: in the last 24 months, how many times has a capital project's underperformance been attributed to a factor outside the project sponsor's span of control?

If that has happened three or more times, the interaction-ownership gap has a pattern, not an exception.

Decision Output

• Decision type: Org design for integration accountability
• Trigger: Any mid-market manufacturer with annual capital plan above 3 to 5 percent of revenue and three or more concurrent capital projects in flight
• Action: Assign one named role (VP of Operational Integration, Director of Capital Planning, or equivalent) with explicit accountability for cross-decision integration. The role sits between Ops, Finance, and HR. Its metric is the gap between approved savings and realized savings.
• Tradeoff: The role costs low six figures fully loaded and requires political cover to work. It can fail if the reporting line is wrong. The role must report to the CFO or COO, not into one of the functions it is integrating across.
• Evidence: Mid-market manufacturers with a named integration owner close 70 to 90 percent of planned savings versus 65 to 80 percent without. The role typically pays for itself in the first capex cycle.

Framework Connection

Nobody Owns the Seams. The phrase is organizational, not operational. The mechanism is a design gap in the accountability structure, not a failure of the people inside it.

This is where the Silo Tax originates. Not in the math. Not in the individual decisions. In the space between them that the organization never assigned an owner to.

The Four Decisions are technically independent as capital lines. They are organizationally independent as reporting relationships. They are not independent on the floor. The third condition is the one the org chart fails to address.

The fix is not a better committee process. It is a named role with explicit accountability for the integration. The model follows the role; the math follows the model.

Strategic Perspective

Org design is the first integration model.

Before a simulation can integrate the Four Decisions, the org chart has to integrate the four owners. Without that, any modeling effort is a side project with no home and no mandate.

For a mid-market CFO, the question is not which simulation tool to adopt or which consulting engagement to run. The question is whose job description includes the sentence: accountable for whether the capital plan delivers as a system, not as a set of projects.

In almost every mid-market manufacturer we work with, that answer is no one. The role is not absent by oversight. It is absent by default org design. Functional orgs optimize for decision clarity, which is why they exist. The Four Decisions are clear. The seams between them are not.

Factories that think assign the seam before they buy the software.