US Peak Surcharges Stack Up: DHL Bills $1.03 a Pound From China, FedEx Hits $117.25

The Analysis

Half past eight on a Thursday morning and a customer who ships accessories out of Shenzhen calls me in the middle of a sentence. He has an eight page quote in front of him, signed off last week, and the invoice that landed in his inbox this morning is eleven hundred dollars heavier than the number his own sales rep agreed to. Before I look at anything I ask him two questions: what is the chargeable weight on that shipment, and what does your tariff say the fuel percentage is applied to. He goes quiet for about four seconds. That silence is this week in one sound.

Here is the pile. DHL is charging one dollar three a pound on international time-definite shipments from mainland China, Hong Kong and Macau to the Americas, running from 1 October 2026 to 5 February 2027, with fuel applied on top of the surcharge rather than beside it. UPS has been staging demand surcharges since 27 September and the top step runs from 22 November to 26 December. FedEx takes additional handling from eight dollars eighty to eleven eighty-five, and oversize from ninety-five seventy-five to one hundred and seventeen twenty-five, both from 23 November. Amazon Shipping starts its three phases on 25 October, a full month before anybody else moves.

Everybody reads those dates. Almost nobody reads the clock. The DHL charge has been running since 1 October, which is twenty three days and five hundred and fifty two hours of freight that went out the door at a price nobody in your building agreed to, and those invoices are already sitting in a queue somewhere. Before we get to what you should do, we are going to walk the timeline hour by hour, from cut-off to delivery, because with this many charges landing on this many different days, the hour you lose is never the hour you were watching.

Start with the meter, because the meter decides who bleeds. Of the four announcements only one is charged on weight. DHL wants a dollar three for every pound of chargeable weight on international time-definite freight out of mainland China, Hong Kong and Macau into the Americas. Everything else in the pile is charged per piece: FedEx wants eleven eighty-five where it used to want eight eighty, and one hundred and seventeen twenty-five where it used to want ninety-five seventy-five. A weight meter and a piece meter behave in opposite ways when you change the size of a box. Make the carton bigger and the piece meter does not care at all. Make the carton bigger and the weight meter invoices you for the air you just bought.

And it is chargeable weight, not actual weight. That one word is where the money goes. If your carton is light and bulky, the volumetric figure runs above the actual figure and the carrier bills the larger of the two, so the surcharge is levied on a number that has more to do with your packaging engineer than with your product. Every square centimetre of dead air in that box is now being billed at a dollar three a pound. I have watched a shipper pay for a carton that was sixty per cent air for two years and never once ask why the freight bill looked the way it did.

Why now, and why these particular windows. Carriers buy peak capacity before they know whether they will need it: extra lift, extra weekend sort shifts, extra seasonal vehicles and extra drivers on the tail. The surcharge is how that money comes back inside the window where the network is genuinely tight. Look at where the windows sit and the reasoning is laid out in the open. UPS tops out from 22 November to 26 December, which is Thanksgiving week straight through to the last shipping day before Christmas. FedEx moves on 23 November, the Monday of Thanksgiving week. Amazon Shipping starts its three phases on 25 October, four weeks ahead of everyone else, which tells you it expects its own network to tighten before the others do.

Then there is DHL, and this is the one I want you to slow down on. Its window runs from 1 October 2026 to 5 February 2027. That is not a Christmas surcharge. Christmas is a rounding error in the middle of it. Five February 2027 is the day before Chinese New Year, which falls on 6 February. So the charge sits across the entire pre-holiday export rush, the fortnight when every factory in the Pearl River Delta and the Yangtze Delta is trying to get finished goods out of the door before the line stops. If your plan was to absorb this through December and be done with it in January, your plan is short by roughly five weeks, and those five weeks are the most expensive weeks of the year to be wrong about.

So who actually bleeds, and by how much. Take a thousand parcels a month that trip additional handling on FedEx. The step from eight eighty to eleven eighty-five is three dollars five a piece, which is three thousand and fifty dollars a month that was not in anybody's budget. Take two hundred oversize pieces a month: the step from ninety-five seventy-five to one hundred and seventeen twenty-five is twenty-one fifty a piece, four thousand three hundred dollars a month. A piece unfortunate enough to trip both charges at once goes from a hundred and four fifty-five to a hundred and twenty-nine ten, up twenty-four fifty-five. Those are piece meters. They punish you for count, not for weight. A shipper of small, dense, well-boxed goods feels those two lines and barely notices the DHL number at all.

The per-pound meter picks a different victim. It goes after whoever ships light and bulky: home goods, pet products, automotive accessories, anything that travels in a big carton with a small item rattling around inside it. That shipper is being charged on a number he can change with a smaller box and a sharper knife, and most of them have never once been told that. I am telling you now, and it is the cheapest money in this whole piece.

Now walk the clock, because the dates in the announcement are not where the damage happens. Take one shipment out of Shenzhen on the evening of Friday 21 November. Cut-off is nominally twenty hundred hours. In practice, on the last working day before a rate change, every shipper in the region wants to be on that night's truck, and the station starts refusing freight from about seventeen hundred because the vehicle goes cube-out rather than weight-out. You have changed nothing about your operation and you have just lost three hours.

Pickup was due at T plus two hours. If the van is full, your cartons stay on the dock and go on the next run, and the next run is tomorrow. There go twenty-four hours, decided in a loading bay at five in the afternoon by somebody who has never seen your invoice. Export sort and linehaul departure sit at T plus eight to fourteen hours in a normal week. In peak the sort plan is capacity constrained, so freight accepted at twenty-one hundred may miss the zero-one-hundred departure and ride the zero-three-hundred instead. Another two to six hours. Not one of these hours appears in any published transit time you have ever been shown.

Then customs on the import side, somewhere around T plus twenty to thirty hours. December inbound volume is the heaviest of the year, and any mismatch between what was declared and what is actually in the carton turns a routine release into a hold measured in days rather than hours. The paperwork side of that is not my desk and I am not going to pretend otherwise. What I can tell you is how it looks from the tracking screen: a scan that stops moving for thirty hours, and a customer service queue that has never been longer than it is in that week.

And then the tail, which is the part I do care about. Last-mile dwell in December runs one to two days longer than it does in October across a lot of US postal codes, and it is not evenly spread: it lands hardest on residential routes and rural ones, which is exactly where your marketplace sellers ship. Add the whole thing up. Three hours at cut-off, twenty-four at pickup, four at sort, thirty at customs, thirty-six on the tail. That is close to four days of drift on a lane you have been promising in five.

So which hour blows first? Not the aircraft. The aircraft was scheduled months ago and mostly runs. The pickup cut-off is the node where every shipper in your building, your industrial park and your entire city competes for the same cubic metre of van space at the same minute. That is the hour worth defending. If you can pull your own internal cut-off forward by four hours and hand cartons to the driver before the station starts refusing, you have bought back more time than any surcharge negotiation will ever buy you.

And your promise breaches earlier than you think. If you have been quoting five business days door to door and you lose the hours I just listed, you are at nine. Marketplace late-shipment metrics do not grade on effort, and a seller account that gets flagged in December does not quietly un-flag itself in January when the network recovers. So the triage question is not which shipment is worth the most money. It is which shipment carries a promise that expires inside the 22 November to 26 December window, because that is precisely where a cost problem turns into a standing problem you will still be dealing with in the first quarter.

Here is the part that is not in the headlines, and it is the reason I am writing this rather than forwarding the announcement. There are two pieces to it. One is that fuel is applied on top of the DHL surcharge, not alongside it. Read that twice. In most express tariffs the fuel percentage is applied to the freight plus surcharges line, which means you are paying a fuel percentage on a peak surcharge. The surcharge is not simply additive; it becomes part of the base that the next multiplier is applied to. On the shipment we are about to price out, that is a hundred and fourteen dollars a shipment. If your particular table applies fuel to base freight only, the figure is zero and you can forget I raised it. Both answers are one email away, and almost nobody is sending that email.

The other piece is the stagger, and it is worse than the compounding. Four carriers, four different dates: 25 October, 23 November, 22 November through 26 December, and a DHL window that runs all the way to 5 February 2027. Because the dates are staggered there is a stretch of about four weeks, from now until the week of 23 November, in which moving freight between carriers or between service levels still saves real money. After that week it does not. FedEx prices up on 23 November, UPS sits at its top step from 22 November, Amazon Shipping has been drawing capacity out of the same pool since 25 October, and the carrier you were going to divert to is now quoting roughly what the carrier you are leaving is quoting. The exit door is not open until 5 February 2027. The exit door closes the week of 23 November. Today is 24 October. That is thirty days, seven hundred and twenty hours, and it is the only clock in this piece you can still do something about.

Let us price one real shipment so the numbers stop floating around. Assume a shipment of accessories from mainland China to the United States, actual weight one hundred and twenty kilos, volumetric weight one hundred and sixty-eight kilos, so the chargeable weight is one hundred and sixty-eight kilos, which is three hundred and seventy point four pounds. Everything after this is an assumption I am stating out loud, not a quote from any carrier. Assume your negotiated base freight is six dollars a kilo purely so the arithmetic has somewhere to stand: one hundred and sixty-eight times six is one thousand and eight dollars.

The DHL surcharge is three hundred and seventy point four pounds times one dollar three, which is three hundred and eighty-one dollars fifty. Assume a fuel rate of thirty per cent. If fuel is applied to base freight plus surcharges, you pay thirty per cent of one thousand three hundred and eighty-nine fifty, which is four hundred and sixteen dollars eighty-five, and one hundred and fourteen dollars forty-five of that is fuel charged on the surcharge. If fuel is applied to base freight only, you pay three hundred and two dollars forty, and the one hundred and fourteen forty-five simply disappears.

Same shipment, one change. Assume you re-box it and get volumetric weight down from one hundred and sixty-eight kilos to one hundred and thirty kilos while actual weight stays where it was, so chargeable weight is now one hundred and thirty kilos, or two hundred and eighty-six point six pounds. The surcharge falls from three hundred and eighty-one fifty to two hundred and ninety-five twenty. That is eighty-six dollars thirty back on a single shipment, plus another twenty-five ninety off the fuel that was riding on top of it. Call it a hundred and twelve dollars a shipment for changing a carton. If you move forty of these a month through the DHL window that is about four and a half thousand dollars a month, and the window runs a little over four months, so call it eighteen thousand dollars across the season.

Now the part that should make you uncomfortable. The DHL charge started on 1 October. Today is 24 October. That is twenty-three days, five hundred and fifty-two hours, and if you are running forty of those shipments a month then roughly thirty-one of them have already gone out of the door with three hundred and eighty-one dollars fifty on each one that nobody in your company had priced. Thirty-one times three hundred and eighty-one fifty is about eleven thousand eight hundred dollars, billed already or in the pipeline already.

Nothing you do from here recovers it, and every day you leave it, another shipment and a half joins the pile. Run it across the full window to 5 February 2027 and you are looking at roughly a hundred and seventy shipments and about sixty-five thousand dollars of surcharge on this one lane. Put that in front of whoever signs off on packaging specifications: it is roughly what a packaging engineer costs for a year, and only one of those two keeps paying you back after February.

Grace Yang, who looks at this from the buying side, would ask what three hundred and eighty-one dollars does to landed cost per unit, and she would be right to ask it: on a unit that sells for forty dollars it is a rounding error per piece, and on a unit that sells for four dollars it is the margin. That is the same triage I have been describing, just entered from the other end of the ledger.

Where could I be wrong? In four places, and you should hold me to every one of them. If your fuel table is base-only, my hundred and fourteen forty-five collapses to nothing and the compounding argument shrinks to a footnote. If your goods are dense and actual weight exceeds volumetric weight on every single SKU you ship, the per-pound meter is far less frightening than the piece meters and your energy belongs on the eleven eighty-five and the one hundred and seventeen twenty-five instead. If your response is to downgrade from time-definite to an economy service to dodge the DHL charge, you save three hundred and eighty-one dollars and buy somewhere between forty-eight and ninety-six extra hours in the worst week of the year, which converts a cost problem into a breach problem; that is a worse trade than it looks on a spreadsheet. And if DHL extends the window past 5 February 2027, which peak surcharges do whenever the network stays tight into the first quarter, then the wait-it-out plan dies and the repack plan becomes the only plan left standing.

None of the work is complicated. It is just time-boxed, and the box shuts in the week of 23 November. Everything I would do falls out of the timeline rather than out of the price list, which is the only reason I have spent this long on the hours instead of on the rates.

Begin by finding out what you have already paid. Pull every quote and every invoice on the China-to-Americas time-definite lane since 1 October and re-price them at a dollar three a pound on chargeable weight. You are not doing this to pick a fight with anybody. You are doing it so that the next thirty days are priced correctly, which is the only part of this you can still influence. It is one morning's work for one person with a spreadsheet, and it is the highest-return morning available to you this month.

Next, get the fuel question answered in writing. One email, one sentence: is the fuel percentage applied to base freight only, or to base freight plus surcharges? Attach your tariff page so the answer comes back tied to a clause. That single answer decides whether your exposure on that lane is three hundred and eighty-one dollars a shipment or something closer to four hundred and ninety-six. Do not accept a verbal answer, because in three months you are going to need to show somebody on paper why the invoice moved.

Then go after the box. Re-measure your top ten SKUs by volume this week, not next month. On the shipment above, cutting chargeable weight from one hundred and sixty-eight kilos to one hundred and thirty kilos is worth a hundred and twelve dollars a shipment, and repacking is the only lever in this whole mess that does not require anybody to agree to anything, does not depend on a carrier conceding a point, and keeps working long after 5 February 2027.

After that, pull volume forward. Anything already packed, invoiced and cleared should be on a truck before the week of 23 November. Aim to get sixty per cent of what you would normally ship in the last week of November out of the door before 22 November. On the DHL lane this does not dodge the surcharge, because that one started on 1 October and you are already paying it; what it dodges is the 22 and 23 November steps and, more importantly, the four days of drift that arrive with them, which is the part that actually costs you the customer.

And rewrite the promise. Every delivery commitment you have published in days needs to become a commitment in hours, with a named daily cut-off and a stated pickup-to-door window, plus twenty-four hours of buffer for Thanksgiving week and forty-eight for the week of 22 December. Test it on twenty live orders and read the scan timestamps rather than the tracking summary. If the timestamps do not match what you promised, change the promise, not the timestamps.

The rest is waiting, and I am not good at waiting. But the sequence matters: find out what you have already paid, get the fuel basis onto paper, cut the air out of the cartons, pull the volume forward, and tell your customers the truth in hours before they work it out in days. Do those five things inside the next seven hundred and twenty hours and the peak stops being something that happens to you. Miss them and you will spend the whole of December doing the thing I hate most, which is explaining to a customer why a parcel that was supposed to be on a porch in Ohio is sitting in a sort facility in Memphis, and that explanation does not get cheaper the more times you give it. Get the parcels out of the door first. Everything after that is accounting.

↑ Back to the key points

— By Nadia Pryce