Why Fixing One Department Rarely Fixes the Result

Four signs the departments are fine and the system is not

In most established companies, the individual functions are run by capable people. Procurement negotiates hard on price. Production keeps its lines utilized. Finance protects the budget. Sales chases the target. Each head of function arrives at the monthly review with green numbers on their own dashboard.

And yet the company-level result—delivery time, cash conversion, margin on new products or customer retention—barely moves. Four patterns usually show up at the same time:

1. Green dashboards, flat outcomes. Each department reports strong performance on its own metric while the customer-facing or P&L metric stays stuck.
2. Improvements that create queues next door. A speed-up or cost reduction in one team turns into waiting stock, rework or overtime in the next team downstream.
3. Decisions that travel up and back down. Anything that crosses two functions has to climb to a shared senior executive before anyone can act, adding weeks of calendar time to a two-hour operational choice.
4. Reorganisations that change nothing. Reporting lines are redrawn every two or three years, but the work on Monday morning follows the same informal path it always did.

Why the "corporate immune system" rejects isolated fixes

When a chief executive brings in a new tool, a lean workshop or an innovation cell inside one department, the rest of the organisation frequently neutralises it within two quarters. Leaders often call this "resistance to change" or a "cultural problem." In almost every engagement we diagnose, the resistance is completely rational.

We call this the corporate immune system—not because people in Procurement, Legal, Finance or Production dislike progress, but because their job descriptions, audit rules and bonus formulas are explicitly designed to protect the existing system against variance:

In a survey of more than 7,600 managers across 262 companies published in Harvard Business Review (2015), Donald Sull, Rebecca Homkes and Charles Sull found that while 84% of managers said they could rely on their own boss and direct reports to deliver most of the time, only 9% said they could rely on colleagues in other functions and units all the time—and just 50% said they could rely on them most of the time. Managers were three times more likely to attribute strategy execution failure to poor coordination across units than to lack of alignment within their own vertical hierarchy. Gartner's May 2024 survey of 1,120 employees similarly reported that only 29% believed their organization helps them navigate cross-functional collaboration effectively, and that employees experiencing high "collaboration drag" were 15 times more likely to feel burned out.

The four domains of an operating-model redesign

Fixing a cross-functional bottleneck does not require a company-wide reorganization. It requires re-architecting the four operating-model connections around the specific workflow that produces the business result:

What this looks like in practice

Across our client work, the biggest operational gains rarely come from making one department work harder inside its own walls. They come from changing the rule, sequence or measure at the boundary between departments:

1. Linking procurement's scorecard to warehouse cost (Client decision)

In Order Backlog Traced to Procurement Incentives, a manufacturer faced a rising order backlog and an overflowing warehouse. The leadership team initially assumed they had a physical capacity problem and prepared a capital proposal to build a new warehouse.

When we traced the physical inventory back to the decision rules that created it, the bottleneck was not in logistics or production. Procurement buyers were measured on unit purchase price. Whenever a supplier offered a volume discount, buyers ordered well beyond the current production plan. Those discounted lots filled the warehouse floor, leaving no space to stage materials for live customer orders. Production stalled waiting for parts that could not be received. Once the company linked warehouse holding cost directly to procurement's performance measure, excess forward buying stopped and the chief executive cancelled the planned warehouse expansion.

2. Uncoupling an inherited serial maintenance schedule (Client-measured)

In Turbine Maintenance with the Resources on Hand, our client—an industrial turbine-maintenance business—wanted to shorten plant overhaul downtime without buying new equipment or hiring additional crews. For years, mechanical, electrical, inspection and safety teams had executed their work in a strict serial handover schedule inherited from historical practice.

Mapping the physical dependencies between tasks revealed that several long steps ran in series only by schedule convention, not because the physical turbine required one task to finish before the next began. Re-sequencing independent tasks to run in parallel using the crews and tools already on site shortened the critical path without capital expenditure (the client now runs the revised sequence; operating effect was not formally measured). We have seen the exact same dynamic on an automotive robotic assembly line, where re-sequencing station handoffs removed the line bottleneck without purchasing a single new robot.

3. Compressing a multi-department store-opening pipeline (Client-measured)

In Store-Development Cycle Compressed Across Six Workstreams, a convenience-store operator's store-opening cycle stretched across site acquisition, legal permitting, store design, contractor tendering, construction and merchandising. Each department optimised its own stage and waited for a 100% sign-off before releasing the file to the next team. By splitting the handoffs so downstream preparation started on bounded early parameters while upstream approvals completed in parallel, the client measured an ~67% shorter opening cycle, holding cost and quality fixed.

Where to start on Monday

Before launching a functional improvement project or redrawing an org chart, run this three-step diagnostic on one stuck business outcome:

1. Pick one end-to-end workflow where every department is green but the P&L result is flat. Good candidates are order-to-delivery, new-product commercialisation, store/branch opening, or capital-equipment turnaround. Do not start with a department; start with the horizontal flow.
2. Put the upstream metric and the downstream cost on the same page. Ask the head of each function in the chain two questions: "What number determines your team's annual rating?" and "When you optimise that number on a bad week, which neighbouring team absorbs the extra cost, inventory or delay?"
3. Change one boundary rule and one shared measure for 90 days. Do not reorganise the reporting lines. Give the two adjacent teams a single joint metric (such as total landed cost of a fulfilled order, or end-to-end cycle time) and a clear decision rule for resolving trade-offs without escalating to the executive committee.

What this looks like in Thailand

In large Thai family-controlled conglomerates and industrial groups, operating-model silos are reinforced by governance history. As Marianne Bertrand, Simon Johnson, Krislert Samphantharak and Antoinette Schoar documented in their study of 93 of Thailand's largest business groups (Journal of Financial Economics, 2008), Thai groups traditionally grew by placing trusted family members or long-serving lieutenants at the head of separate operating companies and functional verticals.

That structure worked exceptionally well during decades of rapid asset expansion: each vertical head had clear personal accountability to the founder or holding board for their unit's budget and headcount. Today, however, when margin growth depends on cross-unit speed—sharing customer data across retail and financial services, coordinating procurement with manufacturing flow, or moving a new product from R&D into commercial scale—that same vertical loyalty acts as a barrier. Middle managers rarely have explicit permission to trade off a portion of their own boss's departmental KPI to improve another unit's result, so every cross-functional friction point waits for the monthly family or executive committee meeting.