DHL Express lifts fuel surcharge for 24-30 August as weekly jet fuel index resets higher

The Analysis

It is 04:20 at the outbound sort and the last wave is still on the belt. The pickup vans went out at 19:00 yesterday, the linehaul is booked for 06:40, and there is a stack of manifests nobody has closed. This is the hour of the week when money leaves a parcel operation without anybody signing for it: a cut-off missed by forty minutes, an address that fails on the first attempt, a remote-area flag nobody cleared before the label printed. None of it looks like a decision. All of it lands on the invoice.

So here is the number for this week. DHL Express posted its fuel surcharge for 24-30 August at 32.25% on US export shipments and 36.00% on US imports. The week before it was 31.75% and 35.50%. Half a point on each leg, and half a point sounds like nothing when you are standing on a dock at four in the morning. Greater China went to 42.5%. Over on the other tables, UPS and FedEx exports were sitting near 47.0% and 46.5%. I want you to hold all five of those numbers, because the distance between them is worth a great deal more than the half-point move that everybody notices.

And there is one line in the DHL table that matters more than any of the percentages. The surcharge is assessed on transport charges plus eligible accessorials. Remote-area fees and residential fees do not simply get added to your bill; they arrive with 32.25% stuck on top of them. Every operational slip that generates an accessorial now pays a 32.25% tax on that slip as well. If you run a parcel desk, that one sentence is what should change your morning.

Let me put a price on the half point first, because if you do not, you will argue about it all week. Assume one US export parcel with a transport charge of 84.00 dollars, a residential delivery fee of 5.20 dollars and a remote-area fee of 12.40 dollars, and assume all three are eligible under your contract. The eligible base is 101.60 dollars. Last week the surcharge on that base was 31.75%, which is 32.26 dollars. This week it is 32.25%, which is 32.77 dollars. The move costs 51 cents a parcel. Assume you tender 9,000 such parcels in the 24-30 August week. That is 4,572 dollars. Real money, but it is not the number I would chase, and it is not a number you can negotiate, because it moves with the index and resets every week. What you can change is the base it is applied to and the table it is applied under.

The bigger lever sits inside the base. Take that same parcel. The remote-area fee of 12.40 dollars does not cost you 12.40 dollars this week. It costs 12.40 plus 32.25% of 12.40, which is 4.00 dollars, so 16.40 dollars. The residential fee of 5.20 dollars costs you 6.88 dollars. Now assume the parcel fails a first delivery attempt and picks up an address-correction fee of 14.00 dollars and a re-attempt fee of 11.00 dollars, both eligible. That is 25.00 dollars of accessorial, and 32.25% on top takes it to 33.06 dollars. The mistake costs 33 dollars, not 25, and that is before you count the van hour, the space it takes on the belt, the second touch by a human being, and the customer who is now writing to you.

Look at the two directions next. 32.25% out and 36.00% in is a gap of 3.75 points. On the same 101.60-dollar base, 36.00% is 36.58 dollars against 32.77 dollars, so 3.81 dollars a parcel. If you are the party named on the inbound waybill, this is your cost and it does not care whether the goods were cheap or expensive. The useful question is which leg you actually control. Where you control the tender, you control the mode, the carton, the accessorials and the carrier. Where you do not, you own the number anyway and the only lever left is how often goods arrive and in what size. I have watched operations spend three weeks arguing about who pays a charge they could have halved by putting two orders in one box.

Then there is 42.5% in Greater China. Against the 32.25% US export rate that is 10.25 points, which on the same base is 10.41 dollars a parcel. Assume you originate 1,200 parcels a week in Greater China and assume your routing and your contract let you inject them at a US gateway instead. The arithmetic says 12,492 dollars a week. I am not going to pretend that is free. Moving an injection point moves your cut-off, your sort window, your linehaul and your exception path, and if your consolidation model depends on holding goods in Greater China you may lose more hours than you save dollars. But 10.25 points is not a rounding error, and if nobody has priced that routing in the last twelve months, the number has been sitting there unchallenged.

The widest spread on the page is between carriers. DHL exports at 32.25% against UPS near 47.0% and FedEx near 46.5%, which is 14.75 and 14.25 points. On the 101.60-dollar base, UPS would bill 47.75 dollars of fuel surcharge and FedEx 47.24 dollars, against DHL at 32.77. That is 14.98 dollars and 14.47 dollars per parcel. Assume you shift 3,000 parcels off UPS onto DHL for this window and you get 44,940 dollars. Shift 3,000 off FedEx and you get 43,410 dollars. Hold one caveat tightly: the eligible base will not be identical across carriers, because each one defines its own base and its own accessorial eligibility. The gap is the point, not the last decimal.

Before anybody gets excited, here is what switching actually costs on the floor. Each carrier runs its own pickup cut-off, its own sort waves, its own linehaul departures and its own exception handling. Assume your current UPS pickup cut-off is 18:30 and DHL wants the parcels at 17:00. That ninety minutes is not paperwork. It is ninety minutes of pick-pack throughput you either have or you do not. If your line runs 220 parcels an hour, ninety minutes is 330 parcels of capacity. If you routinely finish picking at 18:15, moving to DHL pushes 330 parcels into the next sort, and the next sort may not reach the same linehaul. You will have saved 14.98 dollars a parcel on some volume and paid for it with a missed window on other volume. Run that trade in parcels and hours, not in percentage points.

This is why the cut-off clock is the first thing I touch when a table moves. Surcharges reset weekly and I cannot change them. Sort windows open and close every few hours and I can change how I feed them. Take one ordinary parcel. Assume your outbound linehaul departs at 06:40 and your pickup cut-off is 19:00 the previous evening. A parcel labelled at 19:40 misses the van, sits on the dock until the 08:00 pickup, reaches the gateway at 11:20, and misses the 12:00 sort. The next wave runs at 16:30 and the next linehaul after that leaves at 22:10. One forty-minute slip costs somewhere between fourteen and eighteen hours of transit, and in the last mile eighteen hours is the difference between a clean first attempt on Tuesday morning and a failed attempt on Tuesday afternoon with a re-attempt on Wednesday.

Anything that lowers the transport charge also lowers the surcharge, because the surcharge is a percentage of it. Take carton size. Assume 8% of your parcels are shipping in a carton one size too big, and assume right-sizing takes an average of 1.40 dollars off the billable transport charge on each. That is 1.40 dollars off the base, and 32.25% of 1.40 is 0.45 dollars of surcharge you no longer pay. On 9,000 parcels with 720 of them oversized, that is 1,008 dollars a week of base and 324 dollars of surcharge. It is not a big number by itself. It is also thirty minutes of work for whoever owns the carton list, and it is permanent rather than weekly.

Delivery success on the first attempt is the metric that pays back fastest in a week like this, because a failure multiplies. One failed attempt generates at least one accessorial, and that accessorial carries 32.25%. One failed attempt consumes a second van stop, and van stops are the most expensive minutes in the network. One failed attempt puts the parcel back into the sort, where it competes with clean volume for belt space and for the attention of the person who has to decide what to do with it. Assume a first-attempt success rate of 92%, which is 8 failures in every 100 parcels, on 9,000 parcels a week. That is 720 parcels going round a second time.

The fix is mostly a clock. I want an exception queue worked in hours, not in shifts. Assume this rule: any parcel flagged as undeliverable, address-questionable, damaged or held gets touched within four hours of the flag appearing, and gets fixed, re-routed or returned within twenty-four hours. Four hours is not a nicety. Assume a parcel is flagged at 09:00 because the address is missing an apartment number. Catch it inside four hours and you can usually get a corrected address from the customer before the evening sort, and the parcel never leaves the building as an exception. Catch it at 22:00 the next day and the parcel has already failed an attempt, already generated its accessorial plus 32.25%, and is now sitting in a cage with everything else that went wrong.

Re-shipment deserves its own paragraph, because it is where the surcharge quietly doubles. A parcel that fails, comes back to you and goes out again is a new shipment. It pays transport again, it pays its accessorials again, and it pays 32.25% on all of it again. Nothing about the second trip is cheaper because the first one failed. Assume a return-and-re-ship rate of 3% across a 9,000-parcel week. That is 270 parcels paying a second full cycle. At an eligible base of 101.60 dollars each, the second pass carries about 27,432 dollars of eligible charges, of which roughly 8,847 dollars is fuel surcharge. You will never see that as a line on a report. It is buried in outbound volume.

Then there is the part of this that has nothing to do with fuel. Late August is when the ramp starts. Every hour of sort capacity you give away this month is an hour you will not get back in November, because you cannot hire and train a sorter in a week and you cannot add a linehaul that does not exist. Assume your building peaks at 220 parcels an hour on one shift and you are running 165 today. That looks like room. It is not: peak does not arrive evenly, it arrives in two three-hour bursts, and 220 an hour is what your best crew hits on a good afternoon, not what you can hold across a shift. If you are going to test a second carrier, test it now at 165 parcels an hour, not in the week your volume doubles.

One more piece of mechanics. These tables reset weekly. Whatever you decide about 24-30 August is a decision about one week, and the following Monday you start again. That means somebody has to own a weekly routine, and it should not be somebody who also has a dock to run at four in the morning. Assume the routine: Monday 08:00, pull the three published tables, load the percentages into the quoting engine, flag any move bigger than one point, and re-run the top twenty lanes by volume. Twenty minutes. Then go back and audit last month's invoices against the published weekly tables, lane by lane. It is common to find the wrong week's percentage applied to a whole run of shipments. At 32.25% versus 31.75% nobody notices. At 47.0% versus 32.25% somebody should.

Put the whole week on one page, with the assumptions stated. Assume 9,000 US export parcels at an average eligible base of 101.60 dollars including transport, residential and remote-area fees. Total eligible charges 914,400 dollars. The surcharge line at 32.25% is about 294,894 dollars. The half-point move from last week is 4,572 dollars. Cleaning up 150 remote-area flags that should never have been applied, at 16.40 dollars each including the surcharge, is 2,460 dollars. Lifting first-attempt success from 92% to 94% removes about 180 failure events a week; at 33.06 dollars each including surcharge, that is roughly 5,951 dollars. Moving 3,000 parcels off UPS onto DHL is 44,940 dollars. Add them up and you see the shape of the problem: 4,572 dollars is what everybody talks about, and the other three numbers are what nobody does.

The alternatives, and the trap inside each one. Switching carrier: the trap is comparing headline percentages instead of eligible bases, and ignoring that the new carrier's cut-off may cost you ninety minutes of pick-pack. Consolidating: fewer parcels means fewer accessorial events and a smaller base, but it means holding orders, and every hour you hold comes off your delivery promise. Assume a ninety-six hour promise and seventy-two hours of transit, sort and last mile, and you have roughly eighteen hours of hold to spend; spend it in two nine-hour batches rather than one eighteen-hour block, so that a single cancelled linehaul does not take the whole day's volume with it.

Moving injection out of Greater China: 10.25 points is real, but check what it does to your exception path, because a parcel that fails in the wrong place costs more than ten dollars. Repricing the customer: pass the surcharge through as its own line and you will have that conversation with every customer every week; build it into the rate card with a stated review trigger and you have it once a quarter. The trap there is a rate card with no trigger at all, which leaves you eating every move.

Labour is the other half of the cut-off problem, and it is also measured in hours. Assume your sort crew starts at 06:00 and the night wave finishes at 02:00. Assume you move the start to 04:30 for this window. That is ninety minutes of extra sort capacity every morning, which at 220 parcels an hour is 330 more parcels reaching the 06:40 linehaul instead of waiting for the next one. Assume that costs you an extra 1.5 shift-hours a day at a loaded rate you already know. Five days, roughly 7.5 hours of labour against hundreds of parcels that no longer lose fourteen to eighteen hours each. That is the kind of trade I can defend to anybody, because both sides of it are in units the floor understands.

Everything above has been about the export leg at 32.25%. Do not let the import leg sit there untouched at 36.00%. If you receive samples, spare parts, components or replenishment by express into the US, the same logic runs in reverse: the surcharge is assessed on the charges on that inbound shipment, so accessorials on the inbound side are taxed at 36.00%, and every inbound parcel that could have been one parcel but arrived as three has paid the surcharge three times over a larger combined base. Assume you receive 400 inbound express parcels a month and assume a quarter of them could have been consolidated at origin. That is 100 parcels a month, and you do not fix that by calling the carrier. You fix it by telling your supplier how often to ship and in what carton.

One more thing belongs to the parcel desk rather than to procurement: the promise you printed on the product page. Assume your checkout says two to three business days and your real transit, once a parcel misses the 06:40 linehaul, runs closer to fifty hours. Every hour of buffer you quietly spend on a missed sort is an hour taken from a promise somebody else made, and the customer does not experience it as a surcharge problem, they experience it as a late parcel and a message to your support team. Assume each late parcel generates one support contact and each contact costs you twelve minutes of a person. On 720 parcels going round a second time, that is 144 hours of support labour a week. Nobody puts that line next to the fuel surcharge either.

What I would do with the 24-30 August window if it were my dock. I would not spend the week arguing with an index. I would pull the three tables on Monday morning, put the real numbers in the quoting engine, and then spend my hours on the things that actually move parcels: the cut-off clock, address capture at the point of order, the four-hour exception queue, the carton list, and one honest look at which carrier is carrying which lane. The window closes at 06:40 every morning and it does not wait for anybody's approval. Get the parcels out first. Argue about the percentage later.

↑ Back to the key points

— By Nadia Pryce