A stockout is not a supply problem. It is a marketing problem that bills you twice. The first bill is the sales you miss while the offer is unbuyable. The second bill, the expensive one, is the organic rank you give back. Amazon's algorithm rewards listings that sell consistently. Go dark for two weeks and you do not return to your old position. You return to the bottom of the hill you already paid to climb, and you climb it again with ad spend.
Brand owners doing real volume feel this most. Your best ASIN is your best ASIN because it ranks for high-intent terms, converts, and sells every day. That daily velocity is the asset. Run out, and the asset depreciates while a competitor with stock absorbs your impressions and your reviews keep getting older without new sales behind them. Inventory planning is rank protection. Treat it that way and the math changes.
What a stockout actually costs you
When your unit count hits zero, the visible loss is obvious: no sales today. The hidden loss is what matters. Amazon reads a sudden stop in sales velocity as a signal that the listing is no longer the best answer for a query. Your organic placement slides. Sponsored placements stop serving because there is nothing to buy. The Buy Box can drop entirely if you have no offer at all.
Then the offer comes back, and the trouble compounds. You are now ranking lower, so fewer shoppers see you, so you sell less, so the algorithm has less reason to lift you. It is a doom loop, and it runs in the wrong direction. The recovery period usually costs more in ad spend than the safety stock would have cost to hold. You are essentially re-running a launch to win back ground you already owned.
A stockout does not pause your rank. It resets it, and you pay full price to earn it back.
There is a quieter version of this too. Low-stock listings can lose the Buy Box or get throttled before you ever hit zero. If you have ever watched sales soften for no clear reason, inventory is one of the first things to check. It belongs on the same list as the silent revenue leaks like stranded inventory and suppressed offers that drain sales without setting off an alarm.
Forecast from velocity, not from gut
Most stockouts trace back to a forecast built on a feeling. The fix is to forecast from your actual run rate and the things that bend it.
Start with a clean trailing number. Pull your last 30, 60, and 90 days of units sold per ASIN, then look at the trend, not just the average. A product selling 40 units a day and rising needs a different plan than one selling 40 and flat. Use the recent slope to project forward, and adjust for the variables you already know are coming:
- Seasonality. Pull last year's monthly units for the same ASIN. If Q4 ran 2.5 times your summer baseline, your fall reorder has to reflect that, not your current calm.
- Promotions and deals. A Lightning Deal or coupon can triple daily velocity for the duration. Plan the spike and the elevated baseline that often follows.
- Advertising pressure. If you are scaling spend, you are scaling units. Inventory and ad budgets have to move together. When you plan to grow PPC spend as a product matures, the forecast has to keep up with the demand you are paying to create.
- Price changes. A price cut lifts velocity, a hike usually softens it. Bake the expected direction into the number.
The output you want is simple: projected daily velocity for the next 90 days, by ASIN, with the known events marked on the calendar.
Build buffers around lead time, not a round number
A safety buffer is not "keep a few extra weeks on hand." It is a specific calculation built around how long replenishment actually takes and how much that timeline can slip.
Add up your full lead time, end to end: production, freight, customs, and the part everyone forgets, the time inventory sits in Amazon's receiving queue before it is sellable. That last leg has been unpredictable, so treat it as a range, not a fixed number.
Then set a reorder point. The reorder point is the unit count at which you place the next order so you never run dry while you wait. The formula is plain:
Reorder point = (projected daily velocity x total lead time in days) + safety stock
Safety stock is the cushion that absorbs the two things you cannot predict: a demand spike above forecast and a lead time that runs long. Size it to your variability. A steady, predictable seller can run a leaner cushion. A spiky product, or one with a long and shaky supply chain, needs a deeper one. The cost of holding a few extra weeks of a profitable product is almost always smaller than the cost of rebuilding its rank.
Days of cover is your daily dashboard number
Track days of cover for every ASIN: current sellable units divided by projected daily velocity. It converts a raw unit count into the only question that matters, which is how many days until zero. Set thresholds. When a product drops below your reorder point in days of cover, it triggers an order. No debate, no waiting for a monthly review.
Protect rank when stock gets tight anyway
Sometimes the order is late or the spike is bigger than forecast, and you are staring at a few weeks of cover with no resupply in sight. The goal shifts from maximizing sales to stretching the runway so you do not hit zero and lose placement. You sell slower on purpose to keep the listing alive.
Pull the levers that slow velocity without killing the listing:
- Ease off advertising on the at-risk ASIN. Lower bids or pause aggressive campaigns so you stop paying to accelerate toward a stockout. Protect the organic sales, slow the paid ones.
- Nudge price up modestly. A small increase cools demand and protects margin while you wait. Done carefully, this is a controlled slowdown, not a retreat. The same discipline that governs testing an Amazon price increase without tanking rank applies here, just aimed at pacing instead of profit.
- Avoid going fully out of stock at all costs. A live listing with a thin offer holds rank far better than a dead one. Trickling units keeps the velocity signal alive.
This is triage, not strategy. The real win is never needing it, which comes back to forecasting and buffers that gave you enough runway in the first place.
Make inventory part of the whole account
Inventory planning fails when it lives in a spreadsheet that ops owns and nobody else looks at. The reorder decision depends on what advertising is about to do, what creative is about to launch, and what price tests are running. Those teams have to share one calendar. This is exactly why we push brands to stop running PPC, listings, creative, and ops as four separate projects and start running them as one system. The forecast is only as good as the demand signals feeding it, and most of those signals come from the other lanes.
The seasonal version of this is non-negotiable. A major sales event with thin inventory is a guaranteed stockout, and the recovery lands right in your strongest selling window. Inventory is the first item on any pre-Prime Day checklist worth following because it is the one mistake you cannot fix once the surge starts.
Where to start this week
Pick your top five revenue ASINs. For each one, write down four numbers: trailing 30-day daily velocity, total lead time including Amazon receiving, current sellable units, and days of cover. Anything under your full lead time plus a safety cushion is a fire. Order now.
Then set a standing reorder point for each of those ASINs and put days of cover on a dashboard you check weekly, not monthly. That single habit, watching days of cover against a real reorder point, prevents the large majority of rank-killing stockouts. The rank you protect this way is the rank you already paid for. Hold the line on inventory and you stop buying the same ground twice.