Safety Stock

Safety stock is the inventory you hold on purpose, above what your forecast says you need, to absorb variability. Two kinds, specifically: demand that runs hotter than expected, and replenishment that arrives later than promised.

Without it, every forecast miss and every supplier delay becomes a stockout. With too much of it, your working capital sits on a shelf paying storage fees.

Why Stockouts Cost More Than They Look

The obvious cost is the lost sale. The expensive costs are downstream:

• Lost customers, not just orders — the buyer goes to a competitor and may not come back.
• Marketplace ranking damage. Amazon in particular ties sales velocity and availability to placement; a stockout costs ranking that takes weeks to rebuild.
• Wasted ad spend. Campaigns keep running against a page that can't convert.
• Emergency freight. Rush resupply at air-freight prices destroys the margin on the units it saves.
• Broken subscriptions, where a single missed shipment can end a recurring relationship.

That gap between the visible cost and the real one is why safety stock is worth paying for, even though it looks like idle inventory on a balance sheet.

How to Calculate Safety Stock

The most practical formula for ecommerce, because it works from data you already have:

Safety stock = (Max daily sales × Max lead time)
             − (Avg daily sales × Avg lead time)

Worked example:

• Average daily sales: 20 units · Peak daily sales: 45 units
• Average supplier lead time: 30 days · Worst observed lead time: 45 days

(45 × 45) − (20 × 30) = 2,025 − 600 = 1,425 units

That number sizes the buffer against your realistic worst case — a demand spike and a late shipment at the same time, which is exactly when they tend to happen together.

Use your actual observed maximums, not optimistic ones. The supplier's quoted lead time is a marketing number; the one you've actually measured is the planning number.

Safety Stock vs. Reorder Point

They work together and are frequently confused:

• Safety stock — the cushion. You don't plan to sell into it.
• Reorder point — the count that triggers a new purchase order.

Reorder point = (Avg daily sales × Avg lead time) + Safety stock

Using the numbers above: (20 × 30) + 1,425 = 2,025 units. When on-hand hits 2,025, you order — and if everything runs to average, the new shipment lands while the buffer is still untouched.

Setting Buffers by SKU, Not Across the Catalog

A single blanket rule ("hold 30 days of everything") overstocks the tail and understocks the winners. Vary the buffer by:

• Sales velocity. Fast movers justify deeper buffers; slow movers tie up cash. Rank by SKU rather than by product line.
• Lead time and its variability. Overseas manufacturing with unpredictable transit needs far more cushion than a domestic supplier who ships in a week.
• Margin. High-margin items lose more per stockout, so they earn a bigger buffer.
• Shelf life and trend risk. Perishable, seasonal, or trend-driven goods punish overstocking.
• Season. Recalculate before Q4 — both demand variance and supplier lead times get worse exactly when a stockout is most expensive.

What Holding It Actually Costs

Safety stock isn't free, and pricing it honestly is what keeps buffers sane:

• Storage fees, per unit per month, forever.
• Working capital tied up in units that aren't selling yet, at their full landed cost.
• Obsolescence and shrink risk.
• Long-term storage penalties, if it's sitting in FBA.

Storage rates matter more here than sellers expect, because the buffer is the inventory that sits longest. Worth checking against your provider's published storage pricing when you size it — and worth reviewing whenever your lead times change.