Compute a reorder point and treat it as a hard trigger: average daily units multiplied by total lead time in days, plus one production cycle of safety stock. Reorder when units on hand plus units in transit fall below it. Stockouts cost ranking velocity, which is more expensive than the missed sales.
The short version
- The reorder point is arithmetic, not judgment. Daily velocity times total lead time, plus one production cycle.
- Total lead time includes the invisible stages. Inspection, consolidation, customs, drayage, and the check-in queue.
- Sell out and you pay twice. Once in lost orders, again buying back the position you left.
- Size the market before you size the order. Below about $2 million a year, working capital is better deployed elsewhere.
- Throttle advertising when cover gets thin. Buying rank you cannot hold is the most expensive habit in this business.
The number that decides it
Most sellers who run dry did not misjudge demand. They misjudged the calendar.
Take average daily units across the last fourteen days. Multiply by total lead time in days, every stage of it. Add safety stock equal to one full production cycle. The result is the level at which on hand plus in transit triggers your next purchase order.
Worked through: 40 units a day, 25 days production, 7 days inspection and consolidation, 30 days on the water, 10 days customs and drayage, 6 days to check in at the fulfillment center. Total lead time is 78 days. Forty times 78 is 3,120 units. One production cycle of safety stock is another 1,000. Your reorder point is 4,120 units on hand and in transit. Sitting at 3,800 and feeling relaxed means you are already two weeks late.
The stages people forget are customs and check-in, which together often run three weeks. Leave them out and every plan you make is wrong by the exact amount that hurts.
Ten controls, and what each looks like when done properly
- A stage by stage lead time. Done properly means a number in days per stage, confirmed in writing by the factory and the freight forwarder, not an average you remember from last year.
- A validation buy before a scale buy. Done properly means about 200 units at $5,000 to $10,000, with up to four candidate products running at once, before any six figure commitment. You are buying the true return rate and conversion rate, not the units.
- A market big enough to justify the capital. Done properly means a category sized at $2 million a year or more, because below that there is rarely enough revenue to capture profitably once acquisition cost is paid.
- A split first shipment. Done properly means a small air tranche while the container sails, so launch day does not depend on a vessel schedule.
- Buffer stock held outside Amazon. Done properly means a third-party warehouse or your own space, so a restock limit or a receiving delay never becomes a zero.
- Restock limits checked before every order. Done properly means reading your current limits and storage position, because units you cannot send in are capital you cannot use.
- Advertising tied to weeks of cover. Done properly means a written rule, for example throttle broad campaigns under four weeks of cover, rather than a decision made in a panic.
- A weekly reorder review with a name on it. Done properly means one person, one recurring slot, one decision logged. Inventory failures are calendar failures with an absent owner.
- A second qualified factory. Done properly means a supplier that has already produced an approved sample run, not a contact you found last month.
- A written stockout playbook. Done properly means the sequence is decided in advance: raise price to slow the burn, pause discovery campaigns, protect the exact-match terms you rank for.
What running out actually costs
| Cost line | What happens |
|---|---|
| Missed orders | The visible loss, and usually the smallest |
| Ranking decay | Velocity feeds placement, and placement fades while you are absent |
| Advertising history | Campaigns lose the recent performance that earned their placements |
| Competitor gain | Someone else takes the slot, and keeps the reviews that came with it |
| Recovery spend | Rebuying the position costs more per order than holding it would have |
The asymmetry is the whole argument. Carrying an extra month of cover costs you cash and storage fees. Running out costs you cash, position, and the price of buying position back. When the two look close on paper, hold the stock.
What most agencies will not tell you
Inventory is where an agency's incentives and yours quietly separate. A provider paid a share of ad spend has no reason to advise slowing down when cover is thin. A provider on a flat fee has no reason to avoid the conversation. Ask which one you are hiring before you need the advice.
The second thing rarely said out loud: most first order quantities are chosen by the factory's minimum, not by a demand model. A minimum order quantity is your supplier's constraint, not your forecast. When the minimum sits far above what your validation supports, negotiate a smaller run at a higher unit price and book the difference as the cost of information. Paying twenty percent more on 300 units beats owning 3,000 units of a product whose return rate you never measured.
Related answers
- How to source from manufacturers safely and verify quality
- Timeline for Amazon FBA launch from sourcing to PPC
- How to stagger ASIN launches on Amazon
- How to validate a product before Amazon launch
- Amazon launch services: the complete guide
Send us your live velocity and lead times and we will size the reorder point in the free written audit at Flapen.
