What is inventory forecasting?
Inventory forecasting uses historical sales data and external signals to predict how much of each product will sell in a future period. The output of a forecast is not a single number — it is a range, expressed as an expected demand level plus a margin for error. That margin determines how much safety stock to carry.
Forecasting exists to answer one recurring question: how much should I order, and when? Without a forecast, merchants answer this question by guessing — often reordering based on how much they ordered last time, or what intuitively "feels right" for the season. The result is systematic overstocking of slow-movers and chronic understocking of fast-movers.
Demand prediction
A forecast quantifies how much of each SKU will sell in a future period, based on historical data adjusted for trend and seasonality.
Reorder quantity
The forecast directly determines how much to order: enough to cover demand during the supplier's lead time, plus safety stock.
Cash flow planning
A reliable forecast lets merchants plan purchasing spend across weeks and months rather than reacting to stockouts with emergency orders.
Stockout prevention
The primary purpose of forecasting is ensuring stock is ordered before it runs out — not after customers encounter out-of-stock messages.
Why forecasting accuracy matters
Forecast error has a direct financial cost. Over-forecasting leads to excess inventory: capital locked in slow-moving product, warehouse space consumed, and potential spoilage or markdowns. Under-forecasting leads to stockouts: lost sales on the day of the stockout, and sometimes permanent customer attrition if the experience is poor.
The cost asymmetry between these errors varies by product category. For fast-fashion or trend-driven products, overstocking is often more costly than understocking — unsold inventory must be discounted heavily. For replenishment basics or high-margin products with stable demand, the reverse is true: running out of stock costs more than carrying a few extra weeks of inventory.
This asymmetry should directly inform how much safety stock to carry per product. Products where stockout cost exceeds overstock cost should carry more safety stock (higher service level target). Products where overstock cost dominates should carry less.
Inventory forecasting methods
The right forecasting method depends on your data quality, the volatility of demand, and the resources available for analysis. These methods range from simple to sophisticated.
Simple moving average
Calculates the average daily sales over a fixed lookback window (e.g., the last 12 weeks). Easy to understand and implement. Best suited for products with stable, non-trending demand. Weakness: equally weights old and recent data, so it is slow to respond to genuine demand changes.
Weighted moving average
Similar to a simple moving average but assigns higher weight to more recent periods. More responsive to genuine trend changes while still being easy to calculate. Requires choosing weights deliberately — a common approach is 40/30/20/10 for the most recent four periods.
Exponential smoothing
Applies a smoothing factor (alpha, between 0 and 1) to weight the most recent observation most heavily, with weight decaying exponentially for older data. A higher alpha makes the forecast more responsive to recent demand. Most inventory apps use exponential smoothing under the hood for their automated forecasts.
Seasonality adjustment
Calculates a seasonal index for each period of the year (e.g., each week or month) based on how that period's historical sales compare to the annual average. Applies this index to the baseline forecast. Requires at least two years of data for reliable seasonal indices. Essential for any product with predictable seasonal demand patterns.
Trend adjustment
Adds a trend component to the forecast when demand is consistently growing or declining over time. Without trend adjustment, a growing product's forecast will always underpredict, leading to systematic stockouts. Trend is calculated as the slope of demand over a lookback period.
ABC analysis for inventory forecasting
ABC analysis classifies your catalogue into three tiers based on revenue contribution. The purpose is not classification for its own sake — it is to allocate forecasting effort and safety stock investment proportionally to financial impact.
| Tier | Typical share of SKUs | Revenue contribution | Forecasting approach |
|---|---|---|---|
| A | 10–20% | ~70% of revenue | Individual per-SKU forecasting, highest safety stock, weekly review |
| B | 30% | ~20% of revenue | Standard forecasting, moderate safety stock, monthly review |
| C | 50–60% | ~10% of revenue | Simple rules (fixed order quantity or order when low), minimal safety stock |
ABC analysis should be recalculated quarterly. Products that were once A items can become C items as product lines evolve, and new fast-movers should be promoted into the A tier before inventory problems emerge.
Safety stock and economic order quantity
Safety stock is buffer inventory carried above the theoretical reorder point. It exists to cover two types of uncertainty: demand variability (customers buy more than forecast) and supply variability (supplier delivers later than expected). Without safety stock, any positive deviation from expected demand or lead time causes a stockout.
Safety stock formula (practical)
Safety stock = (Max daily sales − Avg daily sales) × Max supplier lead time
Example: if average daily sales are 8 units, maximum daily sales are 12 units, and maximum supplier lead time is 14 days, safety stock = (12 − 8) × 14 = 56 units.
Economic order quantity (EOQ)
EOQ = √(2 × Annual demand × Order cost ÷ Holding cost per unit)
EOQ minimises the combined cost of ordering (fixed administrative cost per PO) and holding (storage, capital, spoilage). In practice, supplier minimum order quantities or volume discount thresholds often override the theoretical EOQ.
Both safety stock and EOQ should be recalculated whenever sales velocity or supplier lead times change significantly. These are inputs to your purchasing decisions, not figures to be set once and forgotten.
Frequently asked questions
More in this guide
Shopify Inventory Management
Complete guide to managing inventory on Shopify
Purchase Orders
Create, send, and track supplier purchase orders
Multi-Location Inventory
Manage stock across warehouses and stores
Inventory Transfers
Move stock between locations accurately
Reorder Points
Automate replenishment with calculated thresholds
Supplier Management
Build and manage vendor relationships
Stocky Alternative
Migrate from retired Stocky to a modern replacement