Little's Law and WIP Limits

Walk into a coffee shop with twenty people inside and two baristas working flat out. Now answer fast: will you get your coffee sooner if more customers crowd in? Of course not — and congratulations, you already understand a sixty-year-old theorem of queueing mathematics. You've just never applied it to your own to-do list. Today that changes.

By the end of this video, you'll be able to predict what happens to waiting time when work-in-progress changes, explain why WIP limits are the single highest-leverage move in this entire course — and you'll have a four-word mantra you will never unlearn.

In 1961, John Little proved a relationship so general it holds for coffee shops, hospitals, factories, and your inbox. In plain terms: lead time equals WIP divided by throughput. How long things take equals how many things are in the system divided by how fast things finish.

Let's run the coffee shop with real numbers. Twenty customers inside; the baristas complete ten orders an hour. Twenty divided by ten: the average customer spends two hours in that shop. Now the manager caps entry at ten customers — same two baristas, same espresso machine, same skill. Ten divided by ten: one hour. Waiting time halved, and nobody worked faster, hired anyone, or bought equipment. The only thing that changed was how much was in progress at once.

That's the whole law, and notice its cruel symmetry: it works in reverse. Let the crowd double and every single customer waits twice as long — including the ones who were already there.

Translate. Every time a team says yes to one more "in progress" item without finishing one first, it silently lengthens the wait for everything already started. Starting feels generous — look how responsive we are! Mathematically, it's the opposite: each new start is a small tax on every unfinished thing. Last lesson's support team, with eleven items in progress and twenty-eight-day lead times? That wasn't a work-ethic problem. It was arithmetic.

And there's a hidden second tax the equation doesn't even show: context-switching. Humans reload slowly. Every jump between tasks burns time on remembering where you were — so high personal WIP doesn't just divide your attention, it shrinks your real throughput too. The equation gets worse in practice than on paper.

Don't take my word for any of this. Run the experiment on yourself this week: pick your five open tasks, and finish them two-at-a-time-maximum instead of all-at-once. Feel what happens to your delivery dates. Better yet — the simulator below this video lets you feel it in ninety seconds.

If lead time equals WIP divided by throughput, and throughput is roughly fixed by your capacity, then WIP is the one lever you control today. A WIP limit is a deliberate cap on concurrent work: "no more than three items in Doing." Simple, visible, brutal — and it transforms behavior, because when the limit is reached, you cannot start. You must finish something first.

That flips the system from push to pull. Push: work is shoved onto people whenever it arrives, regardless of load. Pull: new work enters only when capacity genuinely frees up. The limit is what creates the pull.

And when the limit blocks you and your hands are free? You swarm: go help finish something already in progress. Review a colleague's draft, unblock the stuck item, pair on the hard part. Idle-by-the-limit is not idle — it's the system telling you where your effort actually shortens the queue.

Say the mantra with me, because it compresses this entire lesson into four words: stop starting, start finishing.

You might think starting more things means finishing more things — it's the most natural belief in the working world, and entire careers are built on looking loaded. That's understandable: starting is visible, instant, and feels like progress. Actually, Little's Law says starting more slows everything down while finishing nothing sooner — and context-switching makes it worse than the math. The team that looks slightly under-loaded and finishes relentlessly isn't lazy. It's literate.

Three things to keep. One: lead time equals WIP divided by throughput — more in progress at once means everything takes longer, coffee shops and teams alike. Two: WIP limits are the lever — they force finishing and create pull instead of push; when blocked, swarm. Three: stop starting, start finishing.

Limits need somewhere to live, and they live on a board. But a board is more than three columns and some sticky notes — done right, a board thinks: it confesses where work waits and points at the bottleneck. Next lesson, we make the board honest.