Episode 2 · Theory of Constraints
Find the bottleneck.
You run the pancake breakfast at your kid's school. Last year 240 people came and you fed 120. More ticket sellers and more helpers changed nothing. One griddle decided everything.
Free cheat sheet
Theory of Constraints on one page
How to find the bottleneck, the five steps worked on one griddle, every number from the episode, when it breaks, and two questions to ask this week. Print it or keep it open in a tab. No sign-up.
The idea
Every system has one slowest step, and the whole system can only go as fast as that step. That step is the constraint, or the bottleneck. Eliyahu Goldratt, a physicist, made the idea famous in 1984 with The Goal, a novel about a factory.
His point, in plain words:
An hour lost at the bottleneck is an hour lost for everything. An hour saved anywhere else changes nothing.
In the episode, you time every station on a practice night. Tickets: 180 an hour. Toppings and coffee: 150 an hour. The griddle: 60 plates an hour. So the whole breakfast makes 60 an hour, no matter how many helpers stand at the other tables.
The five steps
- Identify. Look for the pile. Last year it sat in front of the griddle, and just past it, helpers stood idle.
- Exploit. Waste nothing at the bottleneck. A helper took over mixing and plating, so the cook only cooks and the griddle never sits empty.
- Subordinate. Everything else works at the bottleneck's pace. Goldratt called it drum, buffer, rope: the griddle is the drum that sets the beat, a full pitcher of batter is the buffer so it never waits, and the ticket table is the rope, letting people in at the drum's pace.
- Elevate. Add capacity at the bottleneck, and only there. A neighbor lends a second griddle, and herself to run it.
- Repeat. The bottleneck moves. Find the new slowest step and start again.
Worked example from the episode
| Step | Math | Answer |
|---|---|---|
| Last year: people arrive 2 a minute, the griddle does 1. After an hour, 60 are waiting | 60 ÷ 1 a minute | an hour's wait (Little's Law) |
| Before: 4 plates a batch, 3 minutes cooking plus 1 minute with the griddle empty | 4 every 4 min | 60 an hour |
| Exploit: a helper mixes and plates, the cook only cooks | 4 every 3 min | 80 an hour, without buying a thing |
| Elevate: a second griddle | 2 × 80 | 160 an hour |
| Repeat: toppings and coffee are now the slowest step | 150 vs 120 arriving | keeps up |
| Result: 240 came, at $10 a plate | 240 × $10 | $2,400, double last year's $1,200 |
The breakfast is a story. Examples are illustrative numbers, not real data.
When it breaks
- Exploit is not squeeze. The griddle got faster because the cook got help. When your bottleneck is a person, pushing harder is how you break them.
- The bottleneck moves. At 9:30 the coffee urn was empty, and a fresh one takes 10 minutes to brew. The fix was one more buffer: a second urn, always brewing. It gets messy when the bottleneck keeps moving, like in an emergency room.
- Sometimes it's a rule. Sometimes the constraint isn't a step at all, but a rule, like last year's: only the cook touches the batter. Then you change the rule.
- Sometimes it's demand. Once every station beat the 120 people arriving an hour, the constraint left the kitchen. Now it's how many people show up.
Try this week
Pick one process you live with: getting the family out the door, a checkout line, a team's approvals, your inbox. Ask two questions. Where does the pile form? And who is standing around waiting? Then pick one thing to stop wasting at that step. Speeding up any other step changes nothing.
And the next time someone says everyone needs to hurry, ask:
"Where's the bottleneck?"
Sources: Eliyahu M. Goldratt and Jeff Cox, The Goal: A Process of Ongoing Improvement, North River Press, 1984. Eliyahu M. Goldratt and Robert E. Fox, The Race, North River Press, 1986 (drum, buffer, rope). New to the channel? Start with Episode 1, Little's Law.