Business & Organizations · Operations & supply chains

Bottlenecks and the pace of a system

Find the step that limits a production line, see why a desk that never stops working can still fall behind, and use the same thinking to find the tasks that set a project's finish date.

4 short lessons · about 17 minutes · 5 sources · updated November 2026

What you will learn

  • Find the stage that limits capacity in a simple step-by-step process.
  • Tell the difference between a stage being busy and the whole system finishing more work.
  • Calculate how fast a backlog grows when work arrives faster than it can be handled.
  • Separate the time work spends being processed from the time it spends waiting.

Lesson 1 of 4 · 4 min

Why the slowest stage sets the pace

Before you readA workshop cuts 30 toys an hour, paints 20 and packs 25. Every toy needs all three steps. Cutting speeds up to 40 an hour. What happens to finished toys per hour?

Correct.Not quite. The answer is It stays the same.

Finished output stays the same, because painting still handles only 20 an hour. This lesson is about why the slowest required stage sets the pace.

A small workshop cuts, paints and packs wooden toys. Cutting can handle 30 toys an hour, painting 20 and packing 25, and every toy has to go through all three. How many finished toys can the workshop turn out in an hour?

The answer is 20. Painting can't pass along more than 20 an hour, so packing never sees more than that. Painting is the bottleneck, the required stage that limits what the whole line can do.

Eliyahu Goldratt built a management method on this idea, the theory of constraints. He introduced it in 1984 in a novel, The Goal, about a plant manager with three months to save his factory.

Capacity is the rate a system could reach under stated conditions. Actual output can fall short of it. Missing materials, a broken machine, defects or too few orders can all keep the workshop below 20.

The sum also assumes each stage runs at its stated rate and every toy follows the same route. So 20 is a ceiling for this model, and no promise about any given hour.

Now speed cutting up from 30 to 40 toys an hour. Painting still tops out at 20. If cutting runs flat out, 20 unpainted toys pile up every hour, waiting for paint.

Cutting looks more productive, and the workshop finishes toys no faster than before. All it has gained is a bigger pile. That's why judging each department by its own activity can hide what's happening to finished output.

Check yourselfA line cuts 30 units an hour, paints 20 and packs 25, and every unit needs all three steps. In this model, what's the most the line can sustain?

Correct.Not quite. The answer is 20 units an hour.

Every unit has to go through painting, and painting can't do more than 20 an hour, so the line can't sustain more than 20, and other losses could pull it lower. Adding the three rates to get 75 treats the stages as alternatives, but every unit needs all of them. And 25 is packing's capacity, which packing never reaches because painting only sends it 20.

Throughput is the rate at which the whole system turns out usable work. A worker can be busy all day making pieces that sit waiting, need rework or can't reach customers yet. Keeping busy and raising throughput are different things.

To see the difference, managers draw the route. Value-stream mapping charts the flow from suppliers through the factory to customers, which shows where work waits.

Rader Awning, an awning maker in New Mexico, mapped how its orders moved from sales to manufacturing. After the changes that followed, production defects fell 15% and installation corrections fell 25%.

Try to recallA toy line cuts 30 toys an hour, paints 20 and packs 25, and every toy needs all three steps. Why might speeding up cutting leave the number of finished toys unchanged?

Painting is still the required stage that sets the limit. Faster cutting just builds a queue in front of painting, and finished toys come out no faster.

Lessons in this topic

  1. Why the slowest stage sets the pace

    Work out how much a three-step line can finish, and see why a busier stage doesn't always mean more finished work.

    Essentials · 4 min

  2. Why a busy desk still falls behind

    Compare how fast work arrives with how fast it's finished, and see why most of a customer's wait is spent in line.

    Essentials · 4 min

  3. Fix one bottleneck and the limit moves

    Follow the toy workshop after painting speeds up, find where the limit goes next, and see how rework eats capacity without showing in the count.

    Deeper · 4 min

  4. The longest chain of tasks sets the finish date

    Apply bottleneck thinking to a project by finding the critical path and predicting which delays move the finish date.

    Deeper · 5 min

Three quick stories from this topic

The bottleneck novel

The theory of constraints, a way to find bottlenecks, first appeared in a 1984 novel

In The Goal, by Eliyahu Goldratt and Jeff Cox, a plant manager has three months to save his factory. He learns that a line finishes work only as fast as its slowest required stage. In 2011, Time put it among the 25 most influential business management books.

The slowest stage sets the pace

If cutting handles 30 toys an hour, painting 20 and packing 25, the line finishes 20. Speeding up cutting only builds a pile in front of painting.

Then the limit moves

Raise painting to 35 and packing, at 25, becomes the new limit. Goldratt's method ends by sending you back to step one to find the next constraint.

The longest chain sets the date

The fastest a project can finish is the length of its longest chain of dependent tasks

In a textbook schedule for building a house, the longest chain of jobs that must follow one another takes 38 days, so that's the soonest the house can be done. Roofing and appliances sit off that chain, so short delays to them don't move the date.

Why the longest path wins

Every task has to be done, and some can't start until others finish. The project isn't over until its slowest chain of linked tasks is.

Slack

Tasks off the critical path have spare time, called slack. Speeding them up saves nothing. Only shortening a critical task shortens the project.

Nine times the job

In the standard queue model, a 90% busy server means a wait nine times as long as the job

The model is the simplest queue there is, one server with customers arriving at random. Arrivals bunch up, and a server with little idle time can't clear the bunches. At 95% busy, the average wait grows to 19 times the job.

Why spare capacity matters

Idle time is what lets a server catch up after a burst of arrivals. The closer a desk runs to full, the longer each burst takes to clear.

The customer's clock

A request that needs ten minutes of work can take two days to come back. For the customer, the clock runs during the waiting as well as the work.

Sources

This topic is part of From a useful idea to a working business, a guided path through related topics.

  • When variation signals a real problem

    Every process varies from one result to the next. Use a control chart to tell ordinary variation from a real change, and see why adjusting after every result makes a steady process worse.

    4 lessons · 20 min

  • Supply chains

    A cotton T-shirt can cross the Pacific twice before it's worn out. See how goods, orders and money move between the firms in a chain, why lead time sets how much stock you need, and how firms trade low cost against safety.

    4 lessons · 16 min

  • Toyota and lean production

    Toyota learned to build cars with little stock and few defects by letting any worker stop the line. See how just-in-time and built-in quality work, where they came from, and what happened when others copied them.

    4 lessons · 16 min

  • Starting a business

    What it takes to go from a first sale to a firm that lasts. See who pays a new firm's debts, where its first money comes from, who founds businesses and how many are still open five years on.

    4 lessons · 16 min

  • What customer research can and cannot tell you

    Pick research that fits the decision, see why a poll with 2.4 million replies called an election wrong, and tell a rise that follows a change from a rise the change caused.

    4 lessons · 17 min

  • Testing whether a problem is worth solving

    See how Zappos tested its idea with photographs of other shops' shoes, and learn to turn an idea into a problem, an assumption and a small test that could change your mind.

    4 lessons · 17 min

All business & organizations topics · Take the quiz