INSIGHT · SUPPLY CHAIN & OPERATIONS

Bottleneck Management for Machines, Materials and Results

How ABC and LMH, finished goods inventory and contribution margins create a common priority

Ulrich Köster · Interim COO / Supply Chain Executive · 29 September 2026

A factory can be well utilized and still deliver orders late. Raw material sits in the warehouse, but a single purchased component is missing. Meanwhile, finished goods inventory grows for items that nobody currently needs. Each function is working hard, but their priorities are not aligned.

Effective bottleneck management connects machine capacity, material availability, inventory targets and economic order priorities. Its objective is a feasible production plan that meets customer commitments and generates the strongest possible contribution from scarce resources.

The bottleneck machine determines order release

The bottleneck is the resource whose available capacity limits the required throughput in the period under review. Which machine this is depends on the product mix, shift staffing, setup requirements and disruptions. The highest reported utilization alone is insufficient evidence.

Compare the capacity required for the orders actually needed with the capacity realistically available. Also check queues, material shortages and blocking by downstream processes. A persistently full queue can indicate a bottleneck, but it can also result from uncontrolled order release.

The bottleneck receives a reliable sequence and schedule. An appropriately sized buffer protects it from upstream variability. New orders are released according to its consumption and customer due dates. This follows the Drum-Buffer-Rope principle: pace, protective buffer and controlled release. [1]

In practice, this means preparing materials and tools in time, detecting quality problems before the bottleneck wherever possible and deliberately planning setup sequences. The buffer is measured in processable work or time. A pile of incomplete orders protects no machine minutes.

ABC and LMH make material inventory manageable

Raw materials and components need two separate planning perspectives. ABC describes economic significance. Here, LMH describes inventory turnover: Low, Medium and High. Margin and supply criticality are additional decision factors.

Annual consumption value is one suitable ABC basis. A separate ranking by current inventory value shows where capital is tied up. These two ABC analyses must be clearly named: an item can have a high stock value and low annual consumption at the same time.

For LMH, valued consumption in a period is divided by average inventory value in that same period. For finished goods, the corresponding cost of goods sold is used rather than revenue. Periods and valuation must be consistent. L, M and H thresholds are defined by material stage and business model. Movement frequency is supplementary; many transactions do not automatically mean high turnover.

Build the matrix separately for raw materials, purchased components and finished goods, and for relevant sites. Demand variability, lead-time variation, minimum order quantities and shelf life supplement the segmentation. High turnover does not prove stable demand.

A cheap component can block the entire order

A C-item can prevent production of an economically important customer order. Its low purchase price says nothing about the cost of its absence. Add a criticality assessment to the ABC/LMH matrix: which orders depend on it? Are approved alternatives available? When will a replacement actually be ready?

Safety stocks belong where they effectively protect delivery capability or bottleneck performance. An across-the-board stock reduction can remove precisely that protection while leaving the wrong inventory untouched.

Material planning starts with the deliverable order

A material list sorted by shortage quantity does not yet answer what must be procured first. Start with dated customer demand. Bills of material and routings identify the materials that must be available at the right time for the bottleneck and subsequent completion.

Check time-phased available inventory: usable opening stock plus reliably expected, released receipts minus requirements. Reservations must not be deducted twice. Blocked stock counts as available only after a reliable release.

Release an order when its bottleneck slot makes sense and the required materials are secured on time. This does not mean physically stocking every component required later before production begins. What matters is reliable availability at the respective consumption date.

When a shared raw material or component is scarce, allocate it according to feasible order alternatives. Consider incremental contribution per unit of scarce material, committed dates and contractual obligations. A high contribution per finished unit can mislead if it consumes particularly large quantities of the scarce material.

Materials and machines must be feasible together. Material allocation has little value if the machine slot is unavailable. Where several resources are scarce simultaneously, a single ranking is insufficient; an integrated capacity-constrained plan is required.

Finished goods inventory needs a justified purpose

Finished goods can enable short delivery times or cover seasonal demand across a capacity peak. They need a defensible inventory target. Producing solely to maintain utilization moves the problem into the warehouse: scarce machine time has been consumed while the sale remains uncertain.

For stocked finished goods, sales planning, forecast error, replenishment time and service target belong together. Fast-moving standard items may justify a targeted delivery buffer. Slow-moving variants should be assessed for smaller lots, make-to-order or delayed differentiation.

Release for stock must answer two questions: which supply gap does this order prevent? And which customer order would it displace at the bottleneck? An attractive contribution margin on paper does not justify extra production without plausible demand.

Align target inventories across the whole chain. Raw material buffers, WIP and finished goods must not independently protect against the same uncertainty. Inventory reduction and additional contribution are evaluated separately financially; an impairment alone releases no cash.

Align minimum lots for machines and customers

A machine’s minimum production lot and a customer’s minimum order quantity serve different purposes. The production lot determines how much is manufactured after a setup. The customer MOQ determines the quantity accepted per order, line or call-off. They must be aligned but need not be identical.

First distinguish whether there is a technically necessary minimum batch, for example because of a mixing process, tooling, vessel fill or validated process conditions. Or was the lot size chosen solely to spread setup costs over more units? Technical limits must be respected; economic minimum lots need regular review.

A larger lot may save setup time. It also consumes bottleneck capacity and creates additional inventory. Include both the saved setup time and the machine time for quantities not currently required, financing, storage and obsolescence risk. A lower calculated unit cost may accompany a worse overall result.

A fictional example: a customer order requires 200 units, while the internal standard lot is 1,000. At one minute of processing per unit, the additional 800 units consume another 13 hours and 20 minutes at the bottleneck. Whether this advance production makes sense depends on reliable future sales and the orders displaced. The setup benefit alone does not answer the question.

Options include shorter setups, shared campaigns for compatible items or time-bundled demand. A larger production lot can also move downstream in smaller transfer batches, so completed quantities do not wait for the entire lot. Traceability and quality requirements remain mandatory.

The customer order quantity must cover the cost to serve

Customer MOQs should reflect actual service and handling effort: order processing, setup, picking, packaging, transport and use of scarce capacity. Standard goods supplied from shared stock often require a different rule from a customer-specific variant.

Small orders can become economical through transparent small-order surcharges, tiered prices, fixed order windows or consolidated deliveries. Blanket orders with a binding total purchase commitment and smaller call-offs can reduce pressure on production. Define the take-off period, cancellation responsibility, ownership and financing of intermediate stock. A non-binding forecast is not a purchase guarantee.

For release, four quantities must fit: confirmed customer demand, customer MOQ, the technically or economically justified production lot and the permissible target inventory. Supplier MOQs apply upstream as well. Extra quantities resulting from minimum lots must be approved as a deliberate inventory decision, rather than growing unnoticed from the ERP into the warehouse.

Contribution per bottleneck hour changes priorities

Where one machine clearly limits output, relative contribution is a central decision metric. Divide a product’s decision-relevant contribution by its required bottleneck time. A high euro amount per unit can be economically weaker with long machine occupancy than a smaller contribution with short occupancy. [2]

A simplified example with fictional figures: one machine has 80 usable hours. Demand is sufficient for both products, materials are available, and setup differences and other bottlenecks are initially excluded.

Alpha generates more contribution per unit. Beta generates more per bottleneck hour. With 80 hours, 240 units of Alpha would deliver €19,200 contribution, or 480 units of Beta would deliver €26,400. The difference is €7,200 for identical machine time.

This calculation holds only under the stated assumptions. Once Beta demand is fulfilled, remaining capacity is used differently. The real plan must include setup time per lot, scrap, rework, minimum lots and actual material availability. The relevant bottleneck time is that required for saleable good output.

Finance must establish a cost basis appropriate to the decision. Short-term decisions use revenues and costs that actually change with the alternative. Unchanged fixed-cost allocations can distort rankings. Classic throughput accounting uses a narrower measure, revenue less direct material costs, and must not be equated with contribution margin without checking. [2, 3]

Economic priorities need binding constraints

Committed customer dates, contractual duties and strategic minimum supply belong in the plan as constraints. Optimize contribution within the remaining freedom. With a single bottleneck, linear unit values and no additional lot or sequence effects, relative ranking can determine allocation. With multiple bottlenecks or strong setup dependencies, it is only a guide.

Special freight, an additional shift or outsourcing make sense when the genuinely secured incremental contribution and avoided consequential costs justify their extra cost. A generic machine-hour rate alone does not provide a reliable answer.

A practical example with three lines and four customers

The following fictional example connects customer priorities, line economics, material release and lot formation. All quantities, prices and performance figures are assumptions. The planning horizon is two production days, with 16 available hours per line including setup. There is no opening finished goods inventory; all customer orders are due at the end of the horizon.

All three products can be manufactured on every line. Each production lot requires one hour of setup and €40 additional setup cost. The minimum production lot is 300 units; the customer MOQ is also 300 units. This threshold does not require quantities to be multiples of 300. Orders may be split between lines; partial delivery to Customer 4 is agreed.

Good output already includes speed and quality losses. Setup is deducted separately. Line ranking uses the same reference order from Customer 1: (€15 revenue − €6 material and other variable costs − variable manufacturing costs) × good output. This ranks Line 1 before Line 2 before Line 3. Setup and fixed costs have not yet been deducted from this comparative metric.

Customer orders are ranked by margin

The cost basis includes material and other variable costs before line-specific manufacturing. Margin = (price − cost basis) / price. The orders total 4,500 units. Customers 1 and 3 buy the same product P at different prices; their quantities can therefore be manufactured together technically.

This percentage margin is the example’s chosen order priority rule. It differs from contribution per bottleneck hour and does not guarantee the maximum total contribution. The plan makes the effect of a prescribed priority transparent; economic optimization would need to calculate order and line assignments together.

From capacity draft to a material-feasible plan

Customers are initially scheduled in rank order onto Line 1, then Line 2 and finally Line 3. Quantities that no longer fit remain open. This first draft assumes a separate setup for each customer lot. Material availability and minimum lots have not yet been checked, so the draft is not a production release.

The capacity draft contains 3,360 units and six setups. Customer 4 has 1,140 unplanned units. The lots of 200 and 180 units fall below the 300-unit minimum. These deviations must be resolved before release.

A missing material blocks the order

Here, the filter deliberately checks the complete customer order. If all required materials are present, demand is reserved in customer rank order. Shared materials are allocated only once. BOM quantities per good unit include any allowances for material losses. Product Q has no M3 available. Customer 2 is therefore blocked for all 900 units.

Without filling the released slots, the draft initially falls to 2,460 units. Reallocation produces the following: Line 1 manufactures 1,200 P for Customer 1 and 200 P for Customer 3; Line 2 manufactures 400 P for Customer 3 and 720 R for Customer 4; Line 3 manufactures 900 R for Customer 4. Each line occupies exactly 16 hours including setup.

This intermediate plan reaches 3,420 units with five setups. There are 900 open units due to material shortage and 180 due to capacity shortage. The separate 200-unit lot on Line 1 remains below the minimum. The next step therefore combines customer lots of the same product into a shared campaign.

Combined lots make the plan ready for release

On Line 1, the P-orders for Customers 1 and 3 are combined. Since the same product is manufactured without a technical changeover, one setup is sufficient. Customer quantities remain separately traceable in the order allocation.

Line 1: 16 hours − 1 hour setup = 15 production hours.
15 hours × 100 good units per hour = 1,500 good units.

Customer 1 receives its complete 1,200 units; 300 go to Customer 3. Customer 3’s remaining 300 units are manufactured on Line 2. Customer 4 then uses the remaining capacity on Line 2 and Line 3.

Line 2 also reaches exactly 16 hours with two lots: two hours of setup and 14 hours of production. Total output is 3,520 good units with four setups. Every production lot reaches at least 300 units. Combining the lots creates no additional quantity beyond existing customer demand.

The plan consumes 3,520 kg of M1, 1,800 units of M2 and 1,720 units of M4. Remaining stock is 1,480 kg of M1, 200 units of M2 and 280 units of M4; M3 remains at zero. No material is overallocated. The 980 open units are explicitly shown and are initially left unplanned.

The planned contribution after the stated variable manufacturing costs and €160 setup costs is €18,394. This is not net profit; fixed costs have not been deducted. Compared with the intermediate plan that was not yet ready for release, there are 100 additional units and €390 extra contribution: €320 from additional R quantities, €30 from lower-cost P production and €40 from one saved setup.

The management message: material checks determine feasibility; lot formation determines usable capacity. ABC/LMH supports upstream inventory management, but consumption and average inventory data are missing here for a reliable classification of the four materials. If M3 becomes available or a customer date changes, priorities, reservations and line allocation must be reviewed together again.

A common rhythm for Production, Procurement and Sales

Implementation starts with shared decision rules. Planning, Production, Procurement, Sales and Finance must understand the same priority order. Otherwise, each function optimizes its own metric and the bottleneck receives new priorities every day.

Daily S&OE decides on customer orders at risk, material gaps, bottleneck buffers and changes in sequence. Every exception receives an owner and a deadline. An agreed stable planning horizon protects production from unnecessary rescheduling.

Weekly reviews cover bottleneck load, backlogs, shortage causes, setup sequences and ABC/LMH and MOQ exceptions. Monthly S&OP or IBP determines product mix, capacity, inventory targets and supply to economically or strategically important customer segments.

A compact KPI system connects on-time delivery, saleable bottleneck output, hours lost to material shortages, WIP in work hours, finished goods coverage and excess stock. Contributions from actual sales and released working capital are reported separately. This prevents extra warehouse stock from being celebrated as economic success.

Start with one product family

Choose a product family and relevant planning horizon for the pilot. First validate customer demand, BOMs, bottleneck times and available inventory. Then establish ABC/LMH segmentation, criticality attributes and a contribution logic agreed with Finance.

Test order release, material allocation and inventory targets together. Compare on-time delivery, backlog, bottleneck losses, WIP and contribution against a documented baseline. Extend the rules to other product families only once the impact is demonstrable.

The leadership question is: which next order uses our scarce machine and scarce material to fulfil committed customer demand economically? This connection makes bottleneck management a leadership responsibility.

Review your own management logic

The Production & Operations Health Check and the Inventory & Working Capital Health Check on interimcoo.de support a structured initial assessment. They reveal questions about production flow, inventory logic and priorities. Reliable optimization of product mix and improvement potential then requires actual operational data.

Choose the appropriate Digital Solution

Start the free Production & Operations Health Check

Technical foundations

[1] Theory of Constraints Institute · TOC Applications

[2] ACCA · Throughput accounting and the theory of constraints Part 2

[3] ACCA · Relevant costs

Want to assess your own production flow and bottleneck management? Explore the free Production & Operations Health Check to identify operational improvement priorities.

← All English insights · Deutsche Version