# v72 — Product Intelligence

v72 turns Product Master into an operational analytics workspace.

Implemented:
- Product period filter: 7d / 30d / 3m / 6m / 1y / custom.
- Current period vs previous equivalent period.
- KPIs:
  - Net Sales
  - Net Quantity
  - Gross Profit
  - Gross Margin %
  - Average Selling Price
  - Invoice Count
  - Returns / Return Rate
  - Discounts
  - Current Stock
  - Inventory Value
  - Average Daily Sales
  - Days of Inventory
  - Last Sale
  - Last Purchase
- Daily time series: sales / quantity / gross profit / margin / returns.
- Variant analytics including attributes (color / size).
- Branch performance.
- Price history from immutable/effective price entries.
- Purchase cost trend from inventory purchase-receipt ledger.
- Stock Level Over Time reconstructed from immutable inventory movements.
- Day-end stockout rate and explicit methodology note.
- Promotion analysis:
  - before
  - during
  - after
  - quantity lift
  - revenue lift
  - profit lift
  - margin change
  - real promotion usage count when a promotion usage is linked to a sale containing the product
- Customer product analytics:
  - unique customers
  - repeat buyers
  - repeat purchase rate
  - average days between purchases
- Sales by hour / day-of-week API data.
- Return reason/disposition breakdown.
- Explainable Product Health Score.
- Product Intelligence Matrix in Reports Center with:
  - Sales
  - Quantity
  - Profit
  - Margin
  - Discounts
  - Returns
  - Sales trend
  - performance classification.

Important methodology boundaries:
- Stockout currently measures reconstructed day-end stock, not intraday stockout intervals.
- Promotion lift is period correlation, not causal attribution.
- Product Health Score is explainable and is not a financial/accounting truth.
- Reporting aggregates remain the source for product profitability; business ledgers remain source of truth.