Amazon FBA Stockout Prevention: A Seller's Playbook for Never Going to Zero | Inventory Hero
·10 min readStockout Prevention
Amazon FBA Stockout Prevention: A Seller's Playbook for Never Going to Zero
A complete system for preventing Amazon FBA stockouts: how to calculate sales velocity, safety stock, reorder points, and reorder quantities, with the exact formulas and a weekly process you can run today.
For each top-20 SKU, pull the last 14-day and last 90-day average daily units sold from Seller Central or your reporting tool.
2
Compute blended velocity
Compute velocity equal to 0.7 times the 14-day average plus 0.3 times the 90-day average. This weights recent trend without overreacting to a single hot week.
Andrew Erickson is the founder of Inventory Hero. He has spent years working with Amazon FBA sellers on demand forecasting, restock planning, and the cash flow side of running a private-label brand. Inventory Hero exists because every spreadsheet-based inventory system he tried eventually broke — usually right before Q4.
Look up current FBA on-hand units, AWD on-hand units, and units in transit across all suppliers. Add them together for total available.
4
Compute days of cover
Divide total available inventory by blended daily velocity. The result is days of cover at the current sell-through rate.
5
Compare against reorder point
Any SKU within 7 days of its reorder point goes on the order-this-week list. The reorder point is sales velocity times lead time, plus safety stock.
6
Compute reorder quantity and place POs
For each SKU on the list, compute reorder quantity, check it against MOQ and cash on hand, and place the purchase order the same day. Twenty SKUs takes about twenty minutes once the spreadsheet is pre-built.
Frequently Asked Questions
How is reorder point different from safety stock?
Safety stock is the buffer you carry on top of expected demand to absorb variability. Reorder point is the trigger level, the total on-hand-plus-in-transit count at which you place a new PO. Reorder point equals sales velocity times lead time, plus safety stock. Safety stock is a component of reorder point; they are not interchangeable.
What is a good service level for Amazon FBA?
Use a 95 percent service level (z equals 1.65 in the safety stock formula) for best-sellers and core SKUs that drive most of your revenue. Use 80 to 90 percent (z equals 1.28 to 1.65) for long-tail SKUs where occasional stockouts are tolerable. A 99 percent service level (z equals 2.33) is usually overkill on consumer products. The extra inventory cost outweighs the benefit.
How do I calculate sales velocity for seasonal products?
Do not use a single rolling average. Blend a recent average (last 14 to 28 days) with the same period from the prior year, weighted toward whichever is more representative of upcoming demand. For deeply seasonal products like holiday, back-to-school, or summer items, forecast by season and build a separate reorder plan for each peak.
How much inventory should I keep on hand to prevent stockouts?
There is no universal number. The right answer is sales velocity times lead time, plus safety stock, computed per SKU. For most Amazon FBA sellers with 30 to 60 day lead times, this works out to 45 to 90 days of cover for best-sellers and 30 to 60 days for long-tail SKUs. Holding more than 120 days of cover risks long-term storage fees; less than 30 days risks stockouts on any lead-time surprise.
Do AWD and 3PL inventory count toward stockout prevention?
Yes, but only the units that are actually deliverable to FBA within your lead time window. AWD units count toward total available inventory because they can be sent to FBA on demand. 3PL units only count if your 3PL can ship to Amazon quickly enough. Verify the inbound timeline before treating 3PL stock as a buffer.
How quickly does Amazon Best Seller Rank recover after a stockout?
Recovery depends on rank tier and stockout duration. For a top-1000 BSR product after a 7-day stockout, expect 2 to 4 weeks of recovery with aggressive PPC and reorder velocity. After a 14-plus day stockout, recovery is 6 to 10 weeks and may not reach prior rank without launch-style ad spend. The longer the stockout, the more likely a competitor has taken your algorithmic position permanently.
What is the difference between FBA, AWD, and 3PL inventory?
FBA (Fulfillment by Amazon) inventory is in Amazon's fulfillment centers and ships to customers automatically. AWD (Amazon Warehousing and Distribution) is Amazon's bulk storage product, cheaper per cubic foot than FBA, with automatic replenishment to FBA. 3PL (third-party logistics) is non-Amazon warehouse space you control. A stockout-prevention system needs visibility into all three because the unit you do not know about cannot prevent the stockout you are about to have.
A single stockout on a best-selling SKU costs more than most sellers realize. The lost sales are obvious. The damage that compounds, lost Best Seller Rank, lost Buy Box share, lost velocity signal to Amazon's ranking algorithm, takes weeks to recover, and the recovery is asymmetric: you lose rank in days and earn it back in months.
This article is the system most Amazon FBA sellers wish they had been using a year ago. It defines the five numbers that actually prevent stockouts, gives you the exact formulas, walks through a 30-minute weekly process you can run today, and answers the questions most sellers stop and Google halfway through implementing it.
A stockout occurs when a SKU listed on Amazon FBA has zero available units in Amazon's fulfillment network and the listing stops accepting orders. Unlike a temporary "currently unavailable" state caused by an inbound shipment delay, a true stockout means Amazon's algorithm logs zero sales velocity for the SKU, which triggers rank decay and reduced ad eligibility within 24 to 72 hours.
A stockout is not just a missed-sales event. It is an algorithmic event that Amazon's ranking system reacts to, and the reaction lasts well beyond the day inventory comes back in stock.
When a SKU goes to zero on FBA, four things happen, in order, and only the first is obvious.
You lose the sales you would have made. A SKU doing 40 units a day at $32 loses about $17,920 in revenue across a 14-day stockout.
Your Best Seller Rank drops. Amazon's algorithm rewards momentum. When velocity goes to zero, BSR collapses fast and recovers slowly.
Competitors take the rank, the reviews, and the algorithmic real estate. Anyone bidding on your primary keyword sells where you used to sell, and the gap widens daily.
Recovery is asymmetric. Coming back from a stockout typically requires 4 to 8 weeks of intentional ad spend and reorder pressure to restore prior rank.
A useful framing for your own catalog: take your average daily revenue per SKU, multiply by 1.6, and that is roughly the fully-loaded cost of every day you are out of stock once you include rank recovery. That number, not the cost of carrying a little extra stock, is what your inventory system should optimize against.
The math is not hard. The problem is that the math has to be redone, per SKU, every time anything changes. And things change every day.
A typical seller's reorder spreadsheet has columns for current FBA inventory, units in transit, days of supply, supplier lead time, and a "reorder when below X" rule. It works for the first month. Then:
Sales velocity changes seasonally, but the formula is hard-coded against a flat average.
A supplier slips two weeks and the lead time is wrong everywhere.
A new SKU launches and nobody updates its row.
AWD and 3PL inventory live in another sheet, or another person's head.
Urgent SKUs get handled by gut feel, and the spreadsheet becomes decorative.
The realization is not "I need a better spreadsheet." It is "spreadsheets are the wrong tool for a forecasting problem."
A flat 30-day average is too lagging. A 7-day average is too noisy. The right input weights recent days more heavily without overreacting to a single hot week.
For a defensible approximation, blend a 14-day and 90-day average:
Lead time is the entire chain: PO to production, production to port, port to port, port to warehouse, warehouse to FBA receive-and-stow. Most sellers underestimate this by 20 to 40 percent because they only count the manufacturing window.
Track lead time per supplier, per route, and update it with actuals after every shipment lands. If you ship out of Yantian on the same product four times a year, you have enough data to stop guessing.
Safety stock is the buffer inventory held above expected demand to absorb variability. The correct amount of safety stock scales with both the volatility of demand and the length of lead time.
A defensible starting formula:
safety_stock = z × σ_demand × √lead_time_days
Where z is the service-level factor (use 1.65 for a 95% service level on best-sellers, 1.28 for an 80% level on long-tail SKUs) and σ_demand is the standard deviation of daily sales over the last 90 days.
A simpler rule that works for most sellers: hold enough buffer to cover 1.5× your lead time at your peak weekly velocity over the last 12 weeks. Not optimal, but dramatically better than a round number.
A common mistake is ordering by gut feel or by MOQ alone. The right reorder quantity covers your next review period and carries enough to comfortably reach the next planned PO, while respecting your cash position and Amazon's IPI rules.
Sellers tend to think of stockouts and overstock as opposite problems. They are actually the same problem with a different timeline.
Overstock leads to long-term storage fees, IPI penalties, and aging inventory that ties up the cash you need for your next reorder. Six months later, you cannot afford the reorder that prevents the stockout, so you go to zero on a different SKU. Same disease, later diagnosis.
A serious stockout-prevention system buckets inventory by age (Inventory Hero uses 0–90, 91–180, 181–270, 271–365, and 365+ days) so you can run a promotion, bundle, or removal order before slow-moving units start charging you rent.
Manual works while you are under about 25 active SKUs, one marketplace, one supplier per SKU, and steady velocity. Beyond that, three forces compound and the spreadsheet quietly stops being accurate:
More SKUs means more rows that go stale.
Multiple warehouses (FBA, AWD, 3PL) means the on-hand number you trust is wrong because it does not include inventory in transit between your own locations.
Seasonality and promotions mean a flat-average velocity is fighting a moving target.
This is the point where most sellers either hire an operations person to babysit the spreadsheet or move to a forecasting tool. The operations person costs more per month than the software, and they are still using the same flawed inputs.
Sellers who run a disciplined stockout-prevention system typically see:
Stockout days reduced by 40 to 70 percent within 90 days, depending on how chaotic the starting state was.
10 to 50 percent revenue lift on existing SKUs, not from new launches, but from never going low or out on the SKUs that are already selling.
Lower long-term storage fee exposure, because the same system that prevents stockouts flags overstock early.
Hours back every week, because nobody is rebuilding the spreadsheet at 11 p.m. on Sunday.
The point of stockout prevention is not just to avoid the bad days. It is to convert the cash, attention, and rank you currently lose to firefighting into actual growth.
If you take one thing from this guide, make it this: pick your top 5 SKUs by revenue, compute the five numbers above, and place any PO that needs placing today. Then do it again next Monday. The first three weeks are the hardest. After that, the discipline becomes the system.
When you are ready to stop running the spreadsheet, start a free trial and Inventory Hero will do the math for you on every SKU, every day.