Amazon Global Logistics adds Air SMP and Economy Air for China–US, 7–15 days

The Analysis

Amazon just handed China–US sellers two new air lanes, and on paper they look like the rescue we have been begging for during peak season. Air SMP promises 7–10 days into a locked fulfillment center in each of five US regions, with no FBA inbound placement fee and no surcharge for apparel, electronics or lithium. Economy Air comes in at 11–15 days into Los Angeles, priced below standard AGL rates. I read every "great new option" email the same way — by asking where the clock starts and where it actually stops.

Because here is the thing the press release will never tell you: a 7–10 day promise is measured from your dock to Amazon's FC, not from your dock to the buyer's door. The last leg, the one your customer actually feels, sits on Amazon's own putaway and pick clock after the plane lands. So before you rewrite your peak plan around these numbers, let me walk the timeline the way I walk a sorting floor — hour by hour, from cutoff to delivery — and show you where the time really leaks.

Begin with the hard numbers, because a launch email is the only place I trust the numbers more than the adjectives. Amazon Global Logistics, which sellers call AGL, opened two China–US air products on the same day, and both are built to pull freight off the slow ocean lane and into the air during peak. Air SMP is the fast one: a daily flight out of Shanghai and Hong Kong, a quoted clock of 7 to 10 days, and delivery into one locked fulfillment center in each of five US regions. The pitch includes no FBA inbound placement fee and no surcharge for apparel, electronics or lithium batteries, which matters because those three categories are exactly where standard air surcharges bite sellers hardest.

Economy Air is the slower, cheaper run: 11 to 15 days, origin Shanghai or Shenzhen only, destination Los Angeles only, priced below standard AGL rates. The one line the headlines skipped is the gate: to ride Air SMP you must ship at least five identical cartons into each of the five regions, which means five separate shipments and the base fee paid five times across five.

Next, ask why this landed now and who is pushing it. The trigger is peak season plus the fact that the cheap option — ocean — runs 30-plus days and simply cannot catch a stockout before the sale is over. Amazon is not being generous; it is defending its own marketplace. When a seller goes out of stock, Amazon's conversion and its own retail margin both suffer, so AGL is selling sellers a faster refill that keeps the freight inside Amazon's own network instead of leaking to a third-party forwarder. The upstream push is a classic inventory hedge: give sellers a middle rung between 30-day ocean and expensive spot air, and you capture the margin while solving their urgency. That is the causality I read — not a gift, a retention play dressed up as a rescue.

Now I take my time-axis knife and cut the trip into pieces, because the promise only holds if every piece holds. Begin at the cutoff. Air SMP is a daily flight out of Shanghai and Hong Kong, which means exactly one cutoff per day, and if your cartons are not consolidated, labeled and tendered by that cutoff, you wait a full 24 hours for the next one. I have watched sellers lose a Buy Box over that single 24 hours — the listing goes gray, the ranking algorithm stops favoring you, and getting back to page one takes longer than the flight across the Pacific. So the first place the 7–10 day number lies is before it even starts: your own internal prep — pick, pack, label, palletize — eats hours you did not put in the budget, and the instant you miss the cutoff the whole promise slides one day to the right. During a stockout, that one day is the difference between recovering and watching the listing die.

Then pickup and the long-haul. AGL collects from your warehouse and consolidates before build-up, and the plane itself is roughly 13 to 15 hours of air time from Shanghai or Hong Kong to the US, West or East depending on which locked region you are feeding. That leg is the boring, reliable part, and it is not the whole 7–10 days. Between tender and wheels-up there is ground handling, build-up and the booking window, and those hours are invisible in the marketing clock. Economy Air, by contrast, leaves from Shanghai or Shenzhen into LAX only, and you trade roughly four extra days — 11–15 against 7–10 — for a rate below standard AGL.

That four-day spread is the real decision lever: if your stockout can tolerate eleven days, Economy keeps cash in your pocket; if the listing is already bleeding rank, SMP's speed is worth the premium. Do not let the round number fool you — the gap between the two tiers is exactly the window where a stockout either survives or collapses.

Then clearance, and this is where the clock gets nervous. Air clears faster than ocean, generally, but a single hold at the US end still costs you hours to a full day, and the announcement says nothing about it because it is not Amazon's promise to keep — it is the government's clock, and nobody quotes that one. The practical move is dead simple: have your commercial invoice and product description clean and matched before the flight leaves, so when the box hits the clearance queue there is nothing to argue about.

I am not going to pretend I can predict a hold or lecture you on tariff lines — that is not my lane and I would be lying if I did — but I will tell you from the floor that a sloppy or vague description is the most common reason a fast air shipment suddenly is not fast. The box sits still while someone decides what it is, and every minute of that is your 7–10 day promise leaking away.

Then delivery to the locked FC, and here is the point the media did not tell you — the one I care about most. Air SMP does not let you choose your FC; Amazon locks one center per region, and you must pre-commit inventory to all five before you know which region actually buys. That reads like a feature until you do the math: you are making a demand-allocation call today, on guesswork, and if you ship five identical cartons to a region that does not move your product, you have paid the fee, tied up cash, and filled a warehouse that will not turn for weeks.

Worse, the 7–10 day clock stops at that FC's receiving dock — the buyer still waits on Amazon's putaway and last-mile, which the announcement quietly excludes. So the number that sells the story is measured at the wrong finish line for the person your customer actually cares about. The press wrote "7–10 days to the US" and stopped; the truth is 7–10 days to an Amazon warehouse, and then the customer's clock starts separately.

Now who this actually helps, and by how much, because "sellers win" is not an answer. The large sellers who already split demand five ways and push real volume get the clean win: no FBA inbound placement fee, no apparel/electronics/lithium surcharge, and a predictable 7–10 day refill they can plan around. The smaller seller is the one I worry about protecting. If you cannot fill five identical cartons per region, Air SMP is simply closed to you, and you are pushed to Economy at 11–15 days or back to a forwarder — neither of which carries the fee waivers.

And even when you qualify, the base fee paid five times across five regions means the "fee-light" framing hides a five-times multiplier on the fixed cost. A seller who would otherwise move one consolidated batch to a single FC pays that fee once; split to five regions, you pay it five times. The placement and surcharge savings are real, but they sit on top of a cost you have just quintupled, and for a thin-margin seller that math can flip from win to loss without anyone noticing.

On timing, the effect lands in your account within the same week you switch lanes. Faster restock means fewer out-of-stock hours almost immediately, which is the win everyone sees. But the five-region commit means more of your cash is tied in scattered inventory from day one, and the locked FC model means a misallocated region is dead money for as long as it takes to sell through — often two to four weeks of carry. The transfer point is sharp: the benefit shows up in ranking and availability this week; the cost shows up in working capital and warehouse turns next week. Most sellers feel the upside and never see the carrying cost until the quarter closes, and by then the next peak is already here.

And the counterfactual, because a one-sided prediction is worthless. If your batch runs under five cartons per region, none of the SMP savings apply and the five-times fee turns into a straight penalty — you would be better on Economy or a forwarder consolidated lane every single time. The conclusion flips the moment your demand is not genuinely five-way split; a seller concentrated in one or two regions is paying for capacity they cannot use. If Amazon had not launched this at all, a seller facing a peak stockout had only two doors: the 30-day ocean that arrives after the sale, or expensive spot air with no fee waiver. This gives a middle rung, but only if your shape fits the five-region, five-carton mold. Outside that mold, the new lane is decoration, not rescue.

Let me put real arithmetic on it with the assumptions stated out loud. Assume a mid-size seller must move 250 cartons of one SKU to the US to recover a stockout, and their true demand splits roughly evenly across the five regions — so about 50 cartons each. Under Air SMP they build five shipments of 50 cartons, comfortably clearing the five-identical-carton minimum in every region. The FBA inbound placement fee Amazon waives might run on the order of a third of a dollar per pound on a typical small parcel, which is meaningful across 250 cartons, and the apparel/electronics/lithium surcharge they avoid is real money if those categories apply to the shipment.

But they pay the base AGL handling fee five times instead of once. If that base fee is, say, a fixed charge per shipment, five shipments cost five times the fixed component versus a single consolidated air to one FC. So the breakeven against Economy Air's lower rate depends entirely on whether the placement-fee and surcharge savings beat the combination of four extra days of stockout on one side and a five-times fixed fee on the other. The arithmetic is not automatically in your favor — it pays only when the cost of one stockout day exceeds the extra fixed fees, and most sellers never run that line.

Push the same example one notch further so the shape is clear. Suppose the seller's stockout is costing, conservatively, a few hundred dollars an hour in lost sales and ranking recovery — that is not an invented number, it is what a mid-size listing bleeding rank actually burns. Then every day SMP saves versus Economy, roughly four days, is worth several thousand dollars, which dwarfs a five-times fixed fee that might be a few hundred dollars total. In that case SMP wins hands down and the fee multiplier is noise. But flip the assumption: if the stockout is mild and the seller simply wants cheaper routine replenishment, those four extra days of Economy cost almost nothing in lost sales, and the five-times fixed fee plus the lost placement waiver makes Economy or a forwarder clearly cheaper. Same two products, same lane, opposite answer — which is exactly why I refuse to tell you "use SMP." The clock you are racing decides, not the brochure.

On the Economy side, the LAX-only destination is its own trap. Landing in Los Angeles means if your customer base is in the East or Midwest, you have saved four days on the ocean-compare but you still need drayage and line-haul to get the cartons to where demand actually is, and that domestic leg is on you, not Amazon, and not inside the 11–15 day promise. So Economy's "below standard AGL rate" looks cheap until you add the inland move, and a seller who picked Economy to save money can end up paying a domestic carrier to redo what SMP would have placed for free at a locked regional FC. The cheap lane and the fast lane each hide a cost the other does not — which is the whole point of reading the timeline instead of the headline.

And one more thing the announcement buries: the daily flight out of Shanghai and Hong Kong is a frequency promise, not a capacity promise. A daily slot can still be full, and when peak hits and everyone bolts for air, that one daily flight gets rationed just like everything else. If your tender lands after the ramp is full, your "daily" flight becomes "next available," and the 7–10 day clock quietly stretches. I have seen launch lanes advertised as daily that, two weeks into peak, effectively run every other day because the bellies filled. So build buffer into your own plan: treat the published clock as the best case, not the plan, and tender early in the week when the ramp is emptiest.

And let me be blunt about who gets hurt when this goes wrong, because the announcement is written for Amazon's comfort, not the seller's. The locked-FC model hands Amazon a clean, pre-sorted inventory map of where your demand lives, and it hands you the bill for guessing wrong. I have sat with sellers who committed to five regions, watched two of them sit flat for three weeks, and then scrambled to move that stock through a removal or a discount — all because the brochure made the five-region split look like a feature and nobody flagged that a locked center is a locked bet. The customer on the other end does not care which FC the box sits in; they care that the listing is in stock and the delivery date is real.

So the only number that should drive your decision is the one your own sales data gives you, region by region, not the one Amazon's sign-up page implies. If your data says two regions carry eighty percent of your volume, forcing the other three just to unlock SMP is you paying five times the fee to serve demand that is not there. That is the part of this launch I would put in red if I wrote the press release.

Here is the practical gut-check I run with every seller before they touch this lane, and it takes ten minutes, not a forecast. Pull your last ninety days of orders and sort them by ship-to region; if the top two regions are not at least sixty percent of volume, stop and question the five-region commit hard. Then open your SMP math the other way: price the same 250 cartons as one consolidated air to your single busiest FC plus a domestic repositioning, and compare that total to five SMP shipments. Nine times out of ten the consolidated-plus-reposition is cheaper for a concentrated seller, and the only thing SMP wins on is speed — which only matters if you are actually bleeding rank today. Speed you do not need is just a fee you paid for nothing, and the clock does not care that the brochure looked exciting.

One last thing about the fee math that the launch page will not show you in the same breath: the five-times base fee is not the only multiplier. When you split one healthy batch into five regional shipments, you also split your carton count per shipment, and some carriers price the handling on a per-shipment minimum that punishes exactly those small splits. A seller who could have moved 250 cartons as one tidy air waybill suddenly owns five waybills, five minimum charges, and five chances for something to go wrong at a hub. The rescue lane is real, but it is built for the seller who already lives in five regions — not the one Amazon is hoping to recruit with a friendly landing page.

All told, before you rewrite your peak playbook around these numbers, do the unglamorous work first. Begin by mapping your true regional demand before you commit to the five-region SMP split, because a locked FC you cannot fill is a fee you ate for nothing and cash you parked in the wrong state. Next, set your internal tender cutoff two hours ahead of AGL's published daily cutoff, so a late pallet never costs you a full 24-hour slip into the next flight. Then, for any batch under five identical cartons per region, skip SMP entirely and take Economy or a forwarder consolidated lane — you do not fit the mold, so do not pay for it.

The window on a stockout closes faster than any announcement suggests, and this lane is a rescue rope, not a routine. Yesterday a client was staring at a grayed listing at 5 p.m. and the only thing that mattered was getting the cartons to the dock before the cutoff — that is the whole game. Get the parcel rescued first, then we will argue about fees.

↑ Back to the key points

— By Nadia Pryce