Inventory Planning for Marketplaces: Forecasts, Safety Stock and the Cost of Zero

Every stockout is two losses stacked on top of each other. The first is the sale you did not make. The second is the search ranking you built with that sales velocity, which starts decaying the moment your listing shows unavailable. Most sellers plan inventory as if the only risk is buying too much stock. The maths says the risks sit on both sides. The side you ignore is the one that gets you.

A forecast is a discipline, not a guess

A forecast built only on last month’s sales is a guess wearing a spreadsheet. Real demand forecasting starts with a trailing baseline, then adjusts it deliberately. Seasonality moves demand. Sale events move it harder. A planned ad push moves it again. Each adjustment should be a number you wrote down for a reason, not a feeling you had while ordering.

Forecast at the SKU level, never at the account level. A growing account can hide a hero SKU accelerating while half the catalogue quietly dies. The account number looks healthy. The purchase order built on it is wrong in both directions at once.

Then close the loop. Every month, put the forecast next to what actually sold and look at the gap. A forecast that is never compared with actuals never improves. The comparison takes twenty minutes and it is the single cheapest upgrade to your planning.

Safety stock is the price of uncertainty

Safety stock exists because your demand and your supplier are both unreliable in ways you can measure. It is the buffer you hold above expected demand during your replenishment lead time. The logic behind the standard formulas is simple. The more your daily demand swings, the bigger the buffer needs to be. The longer and less predictable your supplier lead time, the bigger again. The higher the availability you want to promise, the bigger still.

The common mistake is applying one buffer rule to the whole catalogue. Your hero SKUs deserve a deep buffer, because a stockout there costs ranking, ad momentum and repeat buyers. Your slowest SKUs cannot afford the same protection, because the holding cost would eat their margin. Decide service levels by tier, on purpose, and write the decision down.

The cost of zero is bigger than the missed sale

Zero stock costs you rankings, ad efficiency and customers, not just revenue. The missed sale is the visible part. The rest compounds quietly.

• Velocity decay. Marketplace ranking rewards recent sales. A stockout interrupts the streak, and the recovery after restocking is a climb, not a snap back.
• Ad waste. Campaigns lose momentum and learning when the listing goes dark, and restarting them costs more than continuing them would have.
• Broken repeat loops. A repeat customer who finds you unavailable buys something else. Some of them like the something else.
• A gift to the competitor. Your stockout is the best marketing your rival never paid for.

The cost of too much is quieter but just as real

Excess inventory does not fail loudly. It bleeds. Every extra unit locks working capital that could be buying more of your winners. Storage charges accrue for as long as the unit sits, and marketplace fee structures generally get harsher as inventory ages. Stock that sits also degrades. Packaging yellows, categories move on, expiry dates approach, and a unit that was sellable in month one becomes a write-off by month ten.

The deepest cost is the one that never appears on an invoice. Cash sitting in slow stock is cash that cannot fund the next purchase order for the SKU that is actually growing. Overstock does not just cost you money. It costs you the growth you would have bought with that money.

When slow stock becomes a liquidation problem

Inventory that has stopped moving is not stock any more. It is a decision you are avoiding. Set an ageing threshold in weeks of cover and treat crossing it as a trigger, not a suggestion. Then walk the ladder in order. Fix the listing if the problem is presentation. Promote if the problem is visibility. Bundle it with a mover if the problem is desirability. Mark it down if the problem is price.

If a SKU survives the whole ladder and still does not move, the honest move is a planned exit. Every month of delay lowers what you recover, because storage costs keep running while the product’s market value keeps falling. An early, deliberate exit recovers more than a late, forced one, and it frees the capital and the storage space for inventory that earns its keep. This is why we treat Inventory Liquidation as a normal part of the planning cycle rather than an admission of failure. The sellers who plan the exit alongside the purchase are the ones who never end up owned by their own stock.