When a product goes out of stock, most sellers calculate the damage as lost sales for the days it was unavailable. Fourteen days out at $800 a day, call it $11,200 of missed revenue, unfortunate but survivable.
That estimate is usually low by a wide margin, because it prices the smallest part of the loss.
The five things you actually lose
The rank position. Organic rank on Amazon is heavily influenced by recent sales velocity. Fourteen days at zero velocity is fourteen days of decay. You do not return to where you were.
The competitor’s gain. Traffic that would have reached you goes somewhere. That competitor gets the sales, and the reviews from those sales, a permanent asset built out of your absence.
The repurchase cost. Coming back in stock means buying rank you already owned, with advertising spend, at current competitive prices. If it took four months of advertising to reach that position originally, rebuilding is neither instant nor free.
Review velocity. No sales means no new reviews. Review recency and velocity feed both conversion rate and rank, so the effect outlasts the outage.
Promotional eligibility. Availability history affects eligibility for Lightning Deals and other placements, sometimes for months after you are back.
The lost margin on the units you could not sell is real. It is rarely the largest line.
Why it keeps happening
It is almost never ignorance of the risk. It is a coordination gap.
The person watching advertising sees improving efficiency and increases budget. The person watching inventory sees a healthy-looking number and does not know a spend increase is coming. Higher spend accelerates velocity. Velocity consumes cover faster than the restock model assumed. The stockout arrives at peak momentum, the worst possible moment, because that is exactly when rank is most valuable.
This is the coordination failure in its most expensive form.
Stock cover is a bidding input
The fix is to treat stock cover as a direct input to bid strategy, not a separate operational concern.
Deep cover: four months or more. Scale. Buy placement, accept a higher ACOS while acquiring rank, push velocity while you can serve it.
Moderate cover: two to four months. Hold. Maintain efficient spend at current levels. Do not start an aggressive push you cannot sustain through the restock cycle.
Thin cover: under six weeks. Protect. Defend branded terms and existing rank. Pull back on discovery spend acquiring new customers you cannot serve. The goal shifts from growth to holding position through the gap.
The specific thresholds depend on lead time and velocity. The point is that the same product warrants materially different bids depending on a variable most PPC managers never look at.
The opposite failure is quieter
Overbuying does not announce itself, and it is still expensive.
Aged inventory surcharges escalate by age band. Long-term storage fees land quarterly and are frequently discovered rather than forecast. Capital sits in a warehouse rather than in the next production run or the next launch.
Inventory age needs tracking continuously so that removal, liquidation or a promotional push is a decision made in advance rather than a reaction to a fee that already posted. The calculation is straightforward: if carrying cost over the next two quarters exceeds recoverable margin at a realistic discount, removal is the cheaper outcome.
Categories with hard ceilings
Some categories cap how aggressively you can ever scale.
Consumables carry expiry dates, which limits how much stock can safely sit in FBA. That ceiling constrains advertising regardless of how well campaigns are performing. It is one of the governing constraints on the food and beverage account in my portfolio.
Gift-led categories concentrate revenue into a few weeks. Restock planning has to be built backwards from those weeks, accounting for FBA receiving times that lengthen precisely when everyone else is shipping too.
A practical planning loop
- Velocity by product, not account average, using recent trend, not a twelve-month mean.
- Cover in weeks, refreshed weekly, factoring in current and planned advertising.
- Lead time end to end: production, freight, customs, FBA receiving, which can add two weeks on its own.
- Reorder point = lead time + safety buffer, sized to demand volatility.
- Bid strategy adjusted to cover band, every week, in the same review.
- Age tracking, so removal decisions happen before the fee arrives.
None of this is sophisticated. It fails through inconsistency rather than complexity, and through being split across two people looking at different screens.