The last job is the one almost nobody does: going back to find out whether any of it worked. It matters for an obvious reason and a less obvious one. The obvious one is that you spent money and effort. The less obvious one is that a change that is never checked reverts — quietly, over about a season, back to whatever the kitchen did before — and nobody notices until the next time somebody stands at the pass with a timer.
📏 Four numbers, twice a year
You already know how to get all of these, because the previous courses built them. Take them before the change and again a month after.
| Number | Where it comes from | The example kitchen |
| Ticket time, median and worst | Twenty tickets at the pass, as in topic 2 | 14 and 26 minutes, now 11 and 17 |
| Tickets with something waiting | The same sheet, the fourth column | 11 of 20, now 3 of 20 |
| Remakes and returns | The waste log from the cost control course | 12 burgers a month, now 4 |
| Hours per station at peak | The roster, from the labour cost course | Unchanged — and that is the point |
Read the last row carefully, because it is the one that gets misused. Nothing here was done to cut hours: the same people worked the same shifts and served the same covers with three fewer minutes on every plate and eight fewer remakes a month. If a change does reduce the hours you genuinely need, that decision belongs to the labour cost course and deserves to be made deliberately rather than as a side effect of buying a basket.
And keep the double-counting rule from the cost control course in mind here too. Those eight remakes a month were already counted as money in that course's waste log. Here they are a signal that the kitchen is calmer, not a second saving to add to the first.
🔁 The re-check, and the honest verdict
Do the twenty-ticket count again a month after any change, and then twice a year regardless. It takes two hours a year in total. What matters is what you do with a bad result, and there are only three honest verdicts.
It worked. Write down what it was, in one line, next to the number. That line is what stops the change being reversed by the next manager who has a different idea.
It worked and moved the jam. The most common outcome, and a good one: the fryer stopped being the bottleneck and now the dish pit is. That is progress, not failure — measure again and start topic 2 from the top, on the new station.
It did not work. Say so out loud, quickly, and find out whether the diagnosis was wrong or the implementation was. Diagnosis wrong means you spent money on a station that was never the constraint; implementation wrong means the change never actually took hold — the new step is not on the sheet, or the machine is being cleaned around rather than used. The second is fixable in a week. Admitting the first one costs some pride and saves the next purchase.
🪦 The equipment graveyard, and how not to add to it
Once a year, walk your own kitchen with fresh eyes and list everything that has not been used in the last month. Every restaurant has some, every item on the list was bought by somebody who was sure, and the reasons repeat with unnerving regularity: it was cleaned for twenty minutes a night until people stopped, nobody owned it, it broke and the part never came, or it was solving a bottleneck that was never measured.
Then do something with the list: sell it, return it, give it away, or put it back into use with a named owner and a line on the closing checklist. Dead equipment is not neutral. It takes space in a kitchen where space is the scarcest thing in the building, and it is a standing argument against the next change you want to make — because everybody remembers the last expensive thing that did not work.
The other habit that keeps the graveyard small is remembering that the analysis expires. A menu change, a volume change or a new channel moves the bottleneck: a restaurant that starts doing serious delivery volume has a completely different peak shape, and the station that was comfortable at a hundred and twenty covers of dining room may not be at ninety plus thirty in bags. When the menu or the volume changes materially, go back to topic 2. It is one Friday with a timer, and it has never once cost more than it saved.
🧭 What you have built
You redefined automation as removing repeated decisions and repeated movements, and adopted the rule that nothing broken gets automated (topic 1). You measured one service and found the actual constraint instead of the loudest one (topic 2). You closed most of the gap for free, by moving work out of the peak, counting steps and using the screen and the sheets you already had (topic 3). You judged a machine by what it removes, who cleans it and what happens when it dies, and sent the money question to the courses that own it (topic 4). You implemented one change, on one station, with a named owner and a rehearsed fallback (topic 5). And you know the four numbers that say whether it stuck.
This opens the operations group, and the next course crosses the pass into the room: light, sound, layout, comfort — the other half of what a guest experiences, and the half that a perfectly running kitchen cannot fix on its own. The finance courses behind you keep doing their job in the background: cost control tells you what a change did to your plate, and the budget tells you every month whether the picture you drew in December is still the one you are living in.