Capacity, Bottlenecks & Constraints

Every process has a speed limit, and it is rarely set by the average step. It is set by the slowest one. That slowest step is the bottleneck, and understanding it is the single most useful idea in operations, because it tells you exactly where to spend your effort — and where extra effort is wasted.

Capacity: how much a step can handle

Capacity is the most a step can produce in a given time — the most cars a paint booth can paint per hour, the most calls a support team can answer per shift. Each step in a process has its own capacity, and they are almost never equal. One station might handle 100 units an hour while the next manages only 40. That mismatch is where the trouble starts, because the steps are connected in a line. Work cannot skip ahead; it has to pass through every station in turn, so a fast station downstream just sits and waits for a slow one upstream to feed it.

The bottleneck sets the pace

Here is the rule that surprises people: a process can only run as fast as its slowest step. If three stations each do 100 units an hour but the fourth manages only 40, the whole line produces 40 an hour — full stop. The fast stations simply pile up work in front of the slow one and sit idle behind it. This slowest step is the bottleneck, named for the neck of a bottle: no matter how wide the body, liquid pours out only as fast as the narrow neck allows.

Constraint: the formal name

The same idea, stated more generally, is a constraint — anything that limits how much the whole system can produce. A bottleneck is usually the constraint, but a constraint can also be a shortage of materials, a limited budget, or a single licensed machine. The practical takeaway is identical: there is almost always one thing holding back the whole, and finding it is the job. In a busy kitchen the constraint might be a single oven; in a support team it might be one specialist everyone has to wait on. Name that one thing and you know where every extra pound of effort will actually pay off.

Why fixing the wrong step wastes effort

The costly mistake is improving a step that is not the bottleneck. Speed up Station A from 100 to 150 units an hour and the line still outputs 40 — you have spent money to pile work in front of the constraint faster. Only by easing the bottleneck itself does total output rise. And there is a twist: relieve one bottleneck and the constraint usually moves to the next-slowest step. Improvement is therefore never “done”; it shifts to wherever the new narrowest point sits.

Improve the bottleneck and the whole system speeds up. Improve anything else and you have spent money for nothing.

Spot the bottleneck situation

Read each scenario and decide whether the bottleneck is real, where it is, or why an improvement won’t help. Tap a card to check.

Sort the improvement decisions

Drag each action into the bucket it belongs to — invest in the bottleneck, don’t waste effort, or prepare for the next constraint. Hit Check placement when you’re done.

How to use it

The skill is resisting the urge to optimise everything and instead asking one question: where is the constraint right now?

• “Don’t speed up packing — the bottleneck is inspection, so that’s where the gain is.” (directs effort to the step that caps output)
• “Our capacity is 40 an hour because Step C can only do 40.” (reads system output from the slowest step)
• “Materials are the constraint this month, not machine time.” (recognises the limit can be off the production line)
• “We eased inspection, so now the bottleneck has moved to shipping.” (expects the constraint to migrate)

Find the one step that sets the pace, fix that, then look again. That loop is most of operations improvement.