Everybody in a restaurant knows where the kitchen is slow, and most of them are wrong. The station that shouts, the person who sweats, the one who complains loudest at the end of service — none of those is evidence. The jam is wherever plates wait, and finding it takes one busy service, a timer and a sheet of paper. It is the single highest-return hour in this course, and it happens before you spend anything.
⏱️ One Friday, twenty tickets
Stand at the pass on your busiest night with a printed sheet and write down four things for twenty tickets: when the ticket arrived, when the last item of it left, how many items it had, and — the important one — whether anything sat finished waiting for something else.
That last column is what turns an anecdote into a diagnosis. A burger going cold under the lamp while its fries finish is not a burger problem; it is a fries problem, and every minute of it is invisible in any report your POS can print. In the example kitchen the twenty tickets gave a median of fourteen minutes and a worst case of twenty-six, and in eleven of the twenty something was sitting waiting. Nine of those eleven were waiting on the same thing.
Two hours of watching, and the conversation stopped being "we get slammed" and became "the fryer cannot keep up between seven and nine". Those are different sentences: the first has no fix, and the second has four.
🧮 Capacity against demand, per hour
Once you suspect a station, the arithmetic is simple enough to do on the same sheet. Work out what that station can produce in an hour, and what the peak asks of it.
The fryer here has one basket and a four-minute cycle, so it produces about 15 portions an hour. On a Friday the restaurant does 120 covers, and half of them arrive in the two peak hours, which is 30 covers an hour; roughly 70% order something fried, so peak demand is about 21 portions an hour.
Fifteen against twenty-one. The gap is six portions an hour, which does not sound like a crisis until you notice it compounds: two hours of running six short is a backlog of twelve portions, and by half past eight every single plate on the pass is waiting on that basket. That is the whole of "we get slammed", written as two numbers.
Do this for each station you suspect — grill pieces per hour against burgers ordered per hour, oven capacity against roasted items, the dish pit against covers. You are looking for the one place where demand crosses capacity first. That is the bottleneck, and it is the only place where spending money changes anything: adding capacity anywhere else just makes plates wait in a different spot.
🕵️ Where the jam usually turns out to be
Before you measure, it is worth knowing the usual suspects, because kitchens are much more alike than they think. In order of how often they turn out to be the real answer:
- One piece of equipment at peak, almost always the fryer, the grill or a single oven — and almost always because it is fine for an average hour and short for the peak one.
- Prep that did not happen, so the line is producing during service what should have existed at four o'clock. This one masquerades as every other problem.
- The expediter, or the absence of one: plates ready at different times with nobody sequencing them, which is why things sit under the lamp.
- Layout, where an ingredient or a tool is not where the hands are, and a station loses seconds hundreds of times a night.
- The dish pit, the most-ignored bottleneck in the industry — a kitchen with no clean sauté pans is a stopped kitchen, and nobody ever writes that down.
- One dish, whose ticket time is double everything else and which drags the whole table with it.
Notice what is not on that list: the number of cooks. Adding people to a kitchen whose bottleneck is a machine or a layout produces a crowded kitchen with the same output, at a permanently higher labour cost — and that cost lives in the labour course, which is where you should go before you hire, not after.
📋 Write down the shape of the jam
Finish the exercise with three lines on paper, because the next three topics all depend on them being specific.
Where: the station and the hours. "The fryer, between seven and nine on Friday and Saturday." Not "the kitchen at the weekend".
How big: the gap in units per hour, or in minutes per ticket. "Six portions an hour short, which is about three minutes added to every affected ticket." A gap you can measure is a gap you can check again afterwards, which is the whole of topic 6.
What it costs: not in pesos — that belongs to the finance courses — but in the things this kitchen can see. Tickets over twenty minutes, plates re-fired because they sat too long, the twelve burgers a month that get remade, the tables that ordered no dessert because they had already been there an hour.
With that written down, topic 3 tries to close the gap for free — and in this restaurant it very nearly does, because six portions an hour is not a machine-sized problem. It only looks like one when nobody has measured it.