What Foresight Needs to Produce Good Forecasts

Foresight learns from data. The quality of predictions depends on the quality and depth of the history available in iVendNext.

4.1 Master data setup

Setup item	Why it matters
Items flagged for forecasting	The sales engine processes only items marked for forecasting. Unflagged items are skipped.
Item lead time	Drives the reorder calculation. Defaults to 7 days when not set.
Default supplier on the item	Required for automatic purchase-order drafting.
Customer territory / segment	Drives churn risk and customer scoring.
Warehouse linked to its company	Needed so every inventory forecast is correctly company-stamped.
Account currency	Ensures the correct currency symbol on every financial forecast.
Company default currency, income account, finance email	Fallbacks for revenue resolution and alert notifications.
Item selling price	Fallback rate for revenue potential when no recent sales exist.

4.2 History depth

Data	Minimum	Recommended
Stock movement	90 days	180 days
Sales	90 days	180 days
Purchases	30 days	90 days
Accounts / ledger	90 days	365 days

Below the minimum, the engine falls back to safe defaults, such as a fixed reorder level and a baseline confidence. Set expectations accordingly. A business that is four weeks live on iVendNext should not be promised high forecast accuracy yet.

4.3 The single most important concept: consumption

Foresight forecasts consumption: the units that leave stock through sales, transfers, and usage. It deliberately ignores stock receipts, because receipts describe supply, not demand. This is why a brand-new item with no sales history cannot be forecast until it has accumulated movement.

Docs

4.1 Master data setup

4.2 History depth

4.3 The single most important concept: consumption