Here's the bottom line up front: for an overseas warehouse, if your ocean transit is 35 days, your safety stock is not 35 days of demand. It's 35 days on the water plus the week your freight spends clearing customs plus the week it takes to get received and put away — and then a buffer for the days your sales refuse to behave. The formula I use on every implementation is simple: stock to hold = daily sales × (transit days + buffer days). What makes it work is filling every number with your own data instead of guessing.
I watched a Shenzhen home-storage seller learn this the expensive way last year. Ten people on the team, one Los Angeles warehouse doing most of their volume. Before peak season they stocked two months of inventory "by feel." One storage box went viral and sold out; another sat there collecting dust. The slow mover ended up occupying half a pallet row while the hot item was out of stock for three weeks. The owner told me he was burning thousands of dollars a day on ads while the product customers wanted wasn't shippable. I've seen this exact mistake three times now, and the root cause is always the same: no formula, just gut feel and courage.
Where Overseas Safety Stock Money Actually Gets Stuck
Most sellers count only the ocean leg. Shenzhen to Los Angeles, 35 days — that number is right. But goods arriving at port are not goods on the shelf. There are three legs between placing an order and having sellable stock, and the ocean is the most predictable of the three.
Leg one is domestic: production and consolidation at the factory, usually 7 to 10 days. When a factory says "three days," believe about half of it — peak-season scheduling and quality rework can each cost you two or three days. Leg two is the 35-day ocean transit, the stable part. Leg three starts at the port: customs clearance, drayage, unloading and putaway at the overseas warehouse. Five to seven days if everything goes smoothly; two weeks if there's congestion or an inspection.
So the honest transit number isn't 35 days, it's 35 + 7 + 7 — roughly 49 days. For a product selling 100 units a day, that's 4,900 units of pipeline and on-hand stock. If you plan on 35 days, the math looks fine on paper, but the shelf is empty during the week you thought the goods had already arrived. Stockouts don't happen at sea. They happen in the week you assumed was already covered.
Count the full cost, too. Money tied up in inventory is three bills at once: the purchase cost, the ocean freight, and the warehouse storage fee. Storage is the one sellers forget to itemize — overseas warehouses bill per pallet or per cubic foot, and a slow mover sitting in a slot is a monthly tuition payment. On the other side, a stockout costs more than a few days of lost sales: your listing rank drops, your ad weight drops, and climbing back after restocking costs more than the sales you lost during the outage. I walk every client through this full accounting before go-live, because only after seeing the total do owners decide between "stock more" and "stock less" on evidence instead of feel.
A Safety Stock Formula You Can Use Tomorrow
On my go-live night checklist, the stocking line has exactly three steps. Skip one and the number is fiction.
Step one: calculate daily sales. Take the last 90 days of real outbound data, strip out the peak promotion days, and average the rest. I take the 90-day average and multiply by a growth factor of 1.1 to 1.2 to leave room for natural growth.
Step two: calculate transit days. Add the three legs: 7 days domestic, 35 days ocean, 7 days clearance and receiving — 49 days total. Adjust it for your own lane; US West Coast is not US East Coast, and a European lane is different again. Write all three legs into a table with an owner and a latest date for each, so nothing hides inside "the shipment is on its way."
Step three: add a volatility buffer. Sales don't move in a straight line, and volatile products need more cushion. My rule: look at the gap between your highest single day and your 90-day average. If the gap is under 30%, buffer 7 days. If it's over 50%, buffer 14 days. Seasonal products — Christmas decor is the classic — need even more, because their spike isn't random, it's concentrated.
Put it together: stock to hold = daily sales × (49 + buffer days). A product selling 100 units a day with a 7-day buffer needs 5,600 units. A niche product selling 20 a day with a 14-day buffer needs 1,260 units. I tell clients to put this table on the replenishment board, one row per product, so the replenishment clerk never has to ask "how much this time" — just follow the table.
| Product type | Daily sales | Transit days | Buffer days | Suggested stock |
|---|---|---|---|---|
| Steady bestseller | 100 units | 49 days | 7 days | 5,600 units |
| High-volatility item | 60 units | 49 days | 14 days | 3,780 units |
| Niche long-tail item | 20 units | 49 days | 14 days | 1,260 units |
| Seasonal product | 80 units | 49 days | 14 days | 5,040 units |
Those are worked examples — plug in your own daily sales. The formula is fixed; your data is what makes it true.
Three Mistakes I've Seen Three Times
Mistake one: counting only the ocean leg. In 2024, an auto-parts client stocked out of a brake pad in their US West Coast warehouse. The vessel had arrived days earlier — the goods sat in the port yard for nine days because a product name on the customs paperwork was wrong, and correcting it took four more days. Thirty-five days at sea, zero problems; thirteen days stuck after arrival. My checklist now has a hard rule: the moment goods leave the factory, send the customs paperwork to the forwarder for pre-review. Don't wait for the vessel to dock.
Mistake two: using peak-season volume as everyday demand. That Shenzhen seller did exactly this — one Black Friday week averaged 400 units a day, so they stocked two months at 400 a day. When the season ended, sales fell back to 120 a day, and the leftover stock sat in the warehouse for four months. I've seen this three times, and every time the call comes three months after peak season: "help us clear inventory." The fix is written above: average 90 days excluding peaks, and run a separate temporary top-up for peak season. Never write peak volume into the standing formula.
Mistake three: one formula for every product. A bestseller and a long-tail item don't fluctuate the same way. Use one buffer for both and you'll either stock out of the bestseller or bury yourself in the long tail. At a pet-supplies client I set up tiers: A-class bestsellers got a 7-day buffer, B-class regulars 10 days, C-class long-tail 14 days but with the order quantity cut in half. Three months later, stockouts fell from five a month to one — and inventory turns actually improved. The counterintuitive truth is this: more stock isn't safer stock; more accurate stock is safer stock. Every extra unit is cash lying in the warehouse; every missing unit is a stockout risk. The formula exists to price both sides, not to stare at only one.
There's one more trade-off most owners don't want to face: safety stock and cash flow pull in opposite directions. The same money turns over six times a year parked in a bestseller and barely once parked in a slow mover. My recommendation is a quarterly product review: any product underperforming its forecast for 90 straight days gets its stock cut in half, and the freed cash goes to A-class items. It's not a glamorous move, but it's the most effective inventory improvement I've seen.
So here's the go-live night checklist: 90-day outbound data pulled; daily sales calculated with peaks stripped out; the three transit legs written into a table; A/B/C tiers decided; customs paperwork pre-reviewed before the goods leave. All five ticked, and your stock number is real. One unticked, and I'd rather ship a week late than place an order on gut feel. A week late costs a week of sales. A wrong guess on gut feel can cost three months of dead stock.
By Lin Xiao, WMS Implementation Consultant








