My data is incomplete.
Most are. The desk shows what it can support and names what is missing beside the measure it blocks, so an incomplete input is visible rather than quietly treated as a zero.
How it works
Four steps, and what to bring to the first one.
Four steps
We identify which exports your systems produce and what context they carry. You do not need every source to begin.
Confirm coverage, item mappings, market comparability and the date each source actually speaks for. Gaps are named, not filled in.
Work through the supported scenarios and item-level choices. Approve what you mean; a suggestion stays a suggestion until you do.
Bring in the next period, see what changed, and revisit the decisions whose evidence moved.
What to bring
To start
Enough to price against and to see what sold.
To go further
Each one unlocks a specific measure rather than a general improvement.
Three practical questions
Most are. The desk shows what it can support and names what is missing beside the measure it blocks, so an incomplete input is visible rather than quietly treated as a zero.
Good. Margin Lab reads supported exports and hands back files. It does not replace a till, a book-keeper or a distributor, and it does not write to any of them.
Weekly or monthly, on the rhythm your data arrives. Each import shows what changed and which approvals need another look.
Explore the sample restaurant, or bring us the decision you are actually weighing.