Working out how many customers you can actually serve

A meal-plan operation has four capacity limits: how much you can cook, how fast you can pack, how many stops you can drive in the window, and how many customer changes one person can process before a cut-off. Most operators plan against the first and are eventually stopped by the third or fourth.

Cooking capacity is the easiest to expand — another shift, another oven, a longer morning. Delivery window capacity is much harder, because it is bounded by traffic and by the hours customers will accept a delivery. Coordination capacity is hardest of all, because it scales with change frequency rather than volume.

That ordering explains why kitchens with spare oven time still feel at their limit. The constraint moved and nobody recalculated.

Twenty stops in one community and twenty stops spread across a city are the same box count and completely different rounds. Delivery capacity is a function of density, not customer count, which is why a service-area decision is really a capacity decision.

It also means growth in an area you already serve is close to free, while growth in a new area carries a step change in cost. Two customers in a distant community may be individually profitable and collectively a loss once you count the drive.

If delivery is the limit, adding kitchen capacity does nothing. The levers are density, window length, or another driver — and another driver is a step cost that only pays back at a certain stop count.

If coordination is the limit, the answer is usually not another coordinator. Self-service pausing, an enforced cut-off and derived production counts each remove work rather than distributing it, and they change the slope instead of the intercept.

The practical discipline is to re-identify the binding constraint every few months. It moves as you grow, and planning against last quarter’s bottleneck is how kitchens invest in the wrong thing.