← All insightsProduction productivity

An hour at the bottleneck is worth more than an hour anywhere else

When production cannot keep up with orders, the first response is often broad: overtime everywhere, a new machine where the noise is loudest, the same efficiency target for every department. The result does not move. The reason is almost always the same: flow through the plant is set by a single step, and until we deal with it, we are dealing with the wrong problem.

A chain is as strong as its weakest link

Every plant has one step with the least capacity relative to what the plan asks of it. It can be a machine, a line, quality inspection, packing, or even one person who prepares set-ups. That step decides how many products leave the plant each day. Everything before it can work faster, but will only pile up work in progress. Everything after it can wait.

Two rules follow, easy to write down and hard to live by:

Why averages hide it

The bottleneck is rarely a secret to people on the shop floor. It is a secret to the reports. Most indicators are built to treat every department the same:

  1. Efficiency is measured the same way everywhere. A department that is not the bottleneck has a 90% utilisation target, so it produces for stock. Work in progress piles up in front of the bottleneck, and cash with it.
  2. OEE is a plant average. An average of 70% does not tell you that one machine runs at 55% and holds up all the others.
  3. Changeovers and start-ups at the bottleneck are treated like anywhere else. One hour of set-up there costs as much as an hour of sales that will not happen.
  4. The plan does not know the bottleneck. Orders are sequenced by date or customer, not by how much time they take on the most loaded step.
  5. Product profitability is measured per unit, not per bottleneck hour. More on this below, because it is the most expensive of the five mistakes.

Which product is really the best

When capacity is limited, what matters is not only how much a product earns, but how much it earns per bottleneck hour. A simple illustration with invented, round numbers:

ProductContribution per unitTime at the bottleneckContribution per bottleneck hour
P€204 min€300
Q€121.5 min€480

Per unit, P looks better. Per bottleneck hour, Q earns 60% more. If sales push P and the plan gives it priority, the company spends its most expensive hour on the less profitable product.

We saw similar logic at a food manufacturer with €17 million revenue. Once product costs were calculated from measured times, the “best” product turned out to be loss-making. The company raised its price, sold fewer units and moved the freed capacity into two more profitable products. Revenue grew by 8% and EBITDA by €219,600 a year.

Five steps, in this order

  1. Find the bottleneck. Where does work in progress pile up? Where is overtime permanent? Which step has the longest queue? Confirm it with measurements, not opinions.
  2. Do not lose time on it. Prepare changeovers in advance, cover breaks, and put quality checks before the bottleneck so it does not process bad parts.
  3. Subordinate everything else to it. Departments upstream produce as much as the bottleneck can consume, not as much as they can make.
  4. Only then add capacity. An extra shift, a second machine or a subcontractor. This is the most expensive step, so it comes last, not first.
  5. Repeat. Once the bottleneck is relieved, it moves elsewhere. Start again there.

Most companies start with step four because it is the most visible. The first three are cheaper and often deliver more.

Five questions for your business

  1. Can your production manager show within a minute which step is the bottleneck today, and back it with a number?
  2. Do you track OEE or utilisation separately for the bottleneck, or only as an average?
  3. Do departments that are not the bottleneck have utilisation targets that make them produce for stock?
  4. Does the production plan take into account the time each order spends at the bottleneck?
  5. Do you know the profitability of your products per bottleneck hour, not just per unit?

If you have no reliable answer to two or more of these, your plant can very likely produce more than it does today, without a new machine.

Where to start

For one week, note where work in progress waits and for how long. For the step with the longest queue, measure how many hours a day it actually produces and where the remaining hours go: changeovers, breakdowns, waiting for material, bad parts. That gap is capacity you already pay for and do not use.

The P and Q table is an illustration with invented numbers, not client data. The food manufacturer’s figures come from case study CS05 and were measured on actual monthly results.

Case study: Production controlling (PDF) ↗

Share this articleLinkedInEmail
Read next
Working capital

Less inventory, more cash, but not more profit

Reducing inventory releases cash, but EBITDA usually falls in the year of the reduction. Three traps, a case from practice, a board table and six steps to keep the warehouse from filling up again.

· 9 min read