Stock that stops being sellable at closing time needs to be counted differently.
Conventional inventory assumes goods persist. You order, you hold, you reorder at a threshold. A tray of croissants does not work like that. It exists for one trading day, its value falls through the afternoon, and at close whatever is left is a loss rather than stock on hand.
So every baked line here carries a same-day quantity, a low threshold, and an aging point measured against the pace that line normally holds. The point is to hear about a problem while it is still solvable.
- Count updates
- On every order
- Thresholds
- Per line
- Aging basis
- That line’s own pace
- Waste recorded
- Per line, per day
What happens to a single line of banana bread between the oven and close.
This is the whole mechanic. Not a stock ledger that gets reconciled weekly, but a count that moves with the till and is judged against how that specific line usually behaves.
- 06:40Tray logged
A quantity is entered when the tray comes out. That number is the day’s ceiling for the line and it only ever falls from here.
- 09:15Selling to pace
Counts fall with each order rung up. The line is tracking against the sell-through curve it normally holds at this hour.
- 13:30Behind pace
Sell-through has fallen below what this line usually does by early afternoon. Flagged now, while there are still four trading hours to act in.
- 16:00Action window
Options offered against the shortfall: bundle it with a drink, feature it at the counter, or mark the remainder down for the last hour.
- 18:30Closed out
Whatever did not sell is recorded as waste against that line, and that record becomes an input to tomorrow’s forecast.
Case / 13:30
The counter-intuitive part
A high remaining count can be the more urgent signal. Three danishes left at half one is a line about to sell out - annoying, but not a loss. Thirteen banana breads left at half one is thirteen units heading for the bin unless something changes in the next few hours.
Counting, alerting, and giving you somewhere to go with the alert.
Live per-item countdown
Each baked line shows units remaining against units baked, updated as orders are taken. Two counters serving at once draw from the same count.
Low-stock alerts you set the level for
A threshold per line, because running out of the signature loaf matters more than running out of the fourth pastry. Alerts reach whoever is on the counter and the owner view.
Aging flags against normal pace
Sell-through is compared to how that line usually moves at that hour on that weekday. Behind pace triggers a flag with hours of runway left.
Multiple bakes per day
A second or third bake of the same line is logged separately, so an afternoon batch is not judged against a six o’clock morning curve.
Discount and bundle prompts
When a line is behind, the system proposes actions with the expected effect on the remainder. Applying one is a decision you make, not something that happens automatically.
Waste recorded honestly
Unsold units are logged per line per day. That number is uncomfortable to look at, which is precisely why it belongs on a report rather than in a bin.
Sell-through history per line
Every line accumulates its own curve. This is what makes the aging flag meaningful and what the nightly forecast reads from.
Ingredient-level depletion
Where a line has a recipe attached, selling it draws down the ingredients behind it, so you see a flour or butter shortfall before prep, not during it.
The last hour is where most avoidable waste is decided.
By the time a café is cashing up, the decision about the remaining stock has already been made by inaction. The window in which something could have been done - a bundle, a counter feature, a staff-suggested add-on, a marked-down last hour - closed somewhere in the afternoon while everyone was busy.
Bake inventory tracking exists to move that decision earlier. The flag fires when the shortfall becomes predictable rather than when it becomes final, and it arrives with concrete options attached rather than as a notification that something is wrong.
What you do with it is a judgement call about your café and your margins. Some owners discount hard at five. Others would rather bundle than mark down, to protect what the item is worth. The system is opinionated about surfacing the problem and deliberately not opinionated about the answer.
How the counting holds up in practice.
What if staff forget to log the morning bake quantity?
How do you handle samples, staff eats and breakages?
Does this replace our supplier ordering?
Our afternoon bake behaves nothing like the morning one. Does that break the pace comparison?
Can we see waste across a whole month?
Daily bake inventory
The first week of counting is what makes every other feature work.
Forecasting, aging flags and waste reporting all read from per-line sell-through history. Getting bake quantities logged is the step that unlocks the rest.
- Your bake lines set up with their own low and aging thresholds
- Counts drawn down by the till, with no separate stocktake to run
- Waste recorded per line so the monthly picture is real