Quick definitions for the metrics and statuses you'll see across Insights, Reports, Inventory Analytics, and the Supply Plan. If a term isn't here, ping us via the chat bubble.

Projected stock curve — For every SKU, Prediko projects your day-by-day stock level: start with what's on hand today, add incoming POs on their expected delivery dates, subtract daily forecasted demand. Most metrics below (days of stock variants, stock at end of month, ordering deadline, recommended order quantity) are just different points read off this same curve. When forecast varies day-to-day (seasonality, promotions, weekends), the curve uses those varying daily values rather than a flat average.

SKUs are coloured by their stock health. For more detail see What is Stock Health and how is it determined?

Stocked out (No Stock) — A SKU with zero stock across your counted warehouses. Virtual warehouses don't count. Red.

At risk — Days of stock left is above zero but at or below your watch-out threshold. Orange.

Healthy — Days of stock left sits between your watch-out threshold and your excess threshold. Green.

Excess — Days of stock left is at or above your excess threshold. More cover than you need — cash tied up. Purple.

Long-term stockout — An active, buyable SKU that has been out of stock for more than 30 days. These usually need a reorder decision or a discontinuation call — they're not noise, they're structural gaps. Shown on the Insights page as Long-term stockouts.

Days of stock left (without incoming) — How many days until you'd run out if no new PO arrives.
Formula: (first stockout day on the no-incoming curve) − today.
With flat demand this collapses to on-hand stock ÷ daily sales, but the curve handles day-to-day variation (seasonality, promotions, weekends) properly.

Days of stock left at delivery — Once your next PO arrives, how many days that delivery (plus any later deliveries) will cover.
Formula: (first stockout day on the with-incoming curve, counted from the PO arrival day onward) − (next PO arrival day).

Stock at end of month (EOM) — Projected on-hand stock on the last day of the month, read directly off the projected curve. Same logic for end of any future month.

Days of stock left at end of month — From the EOM stock level, how many more days until the projection hits zero.
Formula: (first stockout day after end of month) − (last day of month).

Watch-out threshold — How few days of stock you're comfortable with before a SKU is flagged at risk. Configurable per SKU/warehouse or globally. Also referred to as requested days on hand in some surfaces.

Excess threshold — How many days of stock count as too much, before a SKU is flagged excess. Configurable per SKU/warehouse or globally.

Lead time — Days between placing an order and receiving it, set per SKU.

Ending stock — Physical stock on the date in question. We deduct any pre-sold ("sell ahead") units so you don't double-count inventory already promised to customers.

Recommended order quantity — How many units to order now to cover the demand that would otherwise fall on projected stockout days.
Formula: Σ (forecasted daily demand) across every day where the projected curve (with incoming POs) sits at or below zero.

Ordering deadline — Latest date you can place an order without dipping below your watch-out buffer before the next delivery arrives.
Formula: today + (days of stock at delivery − watch-out days).
Place the order after this date and you risk eating into your safety buffer before the next shipment lands.

Forecasted demand (next 90 / 365 days) — Sum of daily forecasted sales over the next 90 (or 365) days.
Formula: Σ (daily forecast) from today through today + 90 (or 365) days. We sum the actual daily forecast values rather than multiplying an average, so seasonality and promotions are properly reflected.

Sales quantity / Units sold — Units sold in the period. Returns reduce this (they're recorded as negative quantities).

Net sales — Total sales revenue in the period. Returns reduce this (negative revenue).

Gross sales — Sales revenue excluding returns (only positive quantities × price).

Bundle quantity — When a bundle SKU sells, the units are decomposed to the child SKUs (parent units × child quantity per bundle). So child SKU reports include their share of bundle sales.

Sell-through rate — The share of available stock that actually sold in a period.
Formula: units sold ÷ (units sold + ending stock) × 100.
Higher = inventory cleared faster. Note this uses ending stock, not units received — it answers "how much of what was around moved?" not "how much of what arrived moved?".

Prediko values inventory using FIFO (first-in-first-out), not weighted average. The cost layers used to compute COGS and inventory value come from your purchase order history.

COGS (Cost of goods sold) — Cost of units sold in the period, FIFO-matched against your purchase order cost layers.

Landed COGS — COGS including landed costs (freight, duties, supplier-side fees). For finished goods, this also includes the proportional landed cost of the raw materials in the BOM.

Unit cost — COGS ÷ units sold for the period. Blank if no sales.

Landed unit cost — Landed COGS ÷ units sold for the period.

Available inventory value — On-hand inventory value at FIFO cost. Excludes any incoming POs.

Landed inventory value — Available inventory value at landed cost. Deliberately excludes incoming POs to avoid double-counting RM landed cost that's already in the finished-goods landed cost.

Incoming inventory (units / value) — Units / value on confirmed POs that haven't fully arrived yet (quantity_confirmed − quantity_received). Shown for today only.

Total inventory value — Available + Incoming. Shown for today only (historical dates show only Available).

Gross margin — Net sales − COGS over the period.

GMROI (Gross margin return on inventory investment) — How much gross margin you earned per unit of inventory tied up.
Formula: gross margin ÷ average inventory cost,
where average inventory cost is the FIFO-valued average of starting and ending inventory across the period.

Planned sales (last 7 days) — Sum of the forecasted sales for the last 7 days from your selected forecasting engine.

Actual sales (last 7 days) — Sum of real sales for the last 7 days from transaction history. Stockout periods are not excluded — if a SKU was stocked out for 3 of those 7 days, those zero-sale days are still in the total.

Forecast accuracy (Plan delta) — How close actuals tracked the plan over the last 7 days.
Formula: (actual ÷ planned) × 100.

This is a simple ratio, not MAPE/WAPE/bias. It tells you direction and rough size of the gap, not statistical forecast quality.

Late PO — A purchase order whose confirmed delivery date is in the past but where units are still in transit (not fully received).

Critical late PO — A late PO with at least one SKU whose current days-on-hand is below 14 days. These are the ones to chase suppliers about today. (Shown as In-transit risk on the Insights page.)