Executive summary
Sales this year
Board expects $54.0M
Operating profit
Board asked for a 10% margin, which is $5.4M
Orders we cannot fill
Sales lost because the factory cannot make the volume, not because demand is missing.
How the year started
above the $7.06M plan
What this means
How this forecast is built, and what it already accounts for
What this forecast already accounts for: summer weather and input costs (sliders on the left), consumer confidence (slider on the left; even an 8-point swing is worth only about $33K, because most of the demand it adds lands in months the factory cannot serve), the Canadian dollar and Mexican peso (priced in, worth less than $0.1M), competitor pricing and US inflation (carried as statistical features because the dataset publishes no elasticity for them), and diesel (sized at about $89K and set aside). Full detail sits in Data, method and audit trail.
What the factory can make, month by month
The grey line is what customers want. The red line is the most the factory can make. Where grey sits above red, those orders are simply lost. The red line moves month to month because monthly capacity is machine-hours, not one fixed number: summer runs all three lines with planned overtime, September gives up 40K litres to scheduled maintenance, and winter runs lighter shifts. And where the plant shipped more last year than its own capacity tab claims, five months of 2025, July by 490K litres, the proven figure is used: a litre that shipped was physically made, so the tab is understated, not the shipment.
The decision in front of the board
Each option is priced as a further step from wherever the sliders currently sit, and shows sales and profit separately, because they do not move together. Adding capacity brings in sales at the cost of making the extra litres; raising price brings in less sales but keeps more of it. Record what was decided in the panel on the left.
Price and cost
What we charge, what it costs, what is left
The gap between the two lines is where all the profit comes from. Everything on this page is the whole year, Q1 actual plus the nine months we are forecasting, and it all moves when you change a setting on the left.
We charge, per litre
It costs us, per litre
We keep, per litre
Gross profit, whole year
Price and cost per litre, three years of history and nine months of forecast
The shaded band is what we keep on every litre. If the gold line rises faster than the black and red lines, the band closes and profit falls even when sales grow.
What is pushing cost
Movement in the three input indices since January 2025.
Could cheaper inputs rescue the year?
If input costs land cheaper or dearer than planned, this is where the year's gross profit lands. Drawn at the temperature and plan settings currently selected on the left.
What the cost is actually made of
Cost per litre is not one number, it is six. Three of them follow a published index and move with the cost slider. The other three are flat rates per litre taken from the Q1 accounts, and they do not move at all.
Which input actually carries the risk
Each bar is the model re-run with that one index moved 10% up and 10% down, everything else held still. Size of spend and size of risk are not the same thing, which is the point of splitting them.
What moves profit most
Each bar is how far operating profit swings between the worst and best case for that one driver, with everything else held still. The longest bar is the thing worth managing first.
Input cost is the longest bar by a wide margin, which is why it has a section of its own. Notice that the weather bar, the thing everyone talks about, is one of the shortest. This chart is measured around the plan of record and does not move with the sliders, because it is the test that was registered before the forecast was run.
The detail
Where the money comes from
One walk from last year to this one, carrying sales, profit and the cost of every ingredient, then the two mix views that explain it. Macro and currency detail lives in Data, method and audit trail.
FY2025 to FY2026, sales and profit in one walk
Each bar shows what that driver did to profit, with what it did to sales underneath. Green adds, red takes away, gold is what the ingredients cost us. Read the gold bars together and you have the whole cost story.
Where the growth is asked to come from
Product mix: a planned split, and one real launch
Monthly P&L: when the money is actually made
A P&L split by channel or product would be invented: the workbook publishes no channel volumes and no costs below the company line. What it can support is sharper. Profit is not spread across the year, it is piled into the summer, in the same weeks the factory runs out.
The risk
What could change it
Could the weather change the answer? Three summers, one ceiling
One question, drawn: could a good summer rescue the year? Revenue adds up left to right, starting from the $7.8M the first quarter already banked, so every line ends at a full-year figure you can compare straight to the tiles at the top of the page. Red is the plan, purple is a heatwave, dark blue is a cool summer, and the gold dashes are the board's $54.0M. Two things to see: the heatwave line barely beats the plan, because the factory ceiling clips every extra litre heat creates; and no line reaches the gold. The downside is real, the upside is capped, and every slider moves this chart.
Five weathers, five P&Ls
The actions
What to do
Four actions, each sized in dollars, with an owner and a date. Edits save in this browser; export to hand the list to the team.
Every action above respects the production ceiling, cold storage, the board's pre-build tolerance, the hedge calendar and contract pricing by channel. The full constraint table, with what each one costs, is in the Constraints tab of Data, method and audit trail.
Decisions on the clock
Every decision ticket applied in Colab lands here automatically with a two-week review window. This list is rebuilt from the decision log on every run, so it cannot go stale or be quietly forgotten.
Action plan
| Action | Financial impact | Owner | Due | Priority | Status |
|---|
Evidence
Data, method and audit trail
Everything behind the numbers, out of the way of the story.
How the data was prepared
What the model learned from, and what proved it
How the baseline is built
Given, derived, or judged
What this data cannot tell us
Model selection
Driver model versus machine learning
Elasticity recovery: what the machine found
The constraints every action respects
Macro and foreign exchange
Currency: CAD and MXN per USD
Confidence and competitor pricing
01_inbox (raw drop, stamped and hashed) → 02_review (open questions) → 03_final (frozen; models read only from here). Only the pipeline moves files and every movement is logged.