Carriers post Sep 7-13 fuel surcharges: FedEx 46%, UPS 46.75%, DHL 43.5%

The Analysis

Yesterday a client messaged me at 21:40 asking why the quote he sent Monday morning did not match the invoice he got Thursday night. Same parcel, same weight, same address. The difference was one number that changed while he was asleep.

For the week of 7 to 13 September, the mainland China fuel tables went to 46.00 percent for FedEx, 46.75 for UPS and 43.50 for DHL. In the same week, DHL South America sits at 33.00 and DHL US-import at 36.75.

He had quoted off last week's table. Not because he was careless. Because nobody had told him the table has a switch-on time, and his quote sheet had a printed number sitting in it with no week attached. This is not a piece about percentages. It is a piece about clocks.

Every surcharge table is a switch, not an average. It goes on at a moment. The week of 7 to 13 September means the number applies from the first minute of the 7th, and the number before it applied until the last minute of the 6th. There is no ramp, no proration, and no grace period that anybody will honour for you at the counter.

Two parcels picked up four hours apart can carry different fuel if the boundary falls between them. That sounds absurd until you have stood at a counter at 23:50 with a manifest that has to close, arguing about which side of midnight a scan belongs to.

The percentage is not the problem. 46.75 is not the problem. The problem is that your quote sheet, your label printer, your pickup route and your customer's promise all run on different clocks, and nobody in the building owns the gap between them.

Four clocks matter and they are almost never synchronised. The carrier's table flips at the week boundary. Your quote template carries whatever validity someone typed into it, usually one week, sometimes a month. Your pickup clock is a driver's route, which is not a clock at all. Your customer's clock is the moment they promised their own customer, which is almost always earlier than the other three.

The gap between the carrier clock and your quote clock is where the money leaks. Assume your template says valid for one week. Assume you build the quote Sunday night. At 00:00 Monday the table moves, and your quote is stale before the customer has opened it. You have committed to a number you cannot buy.

Change the validity rule and most of this goes away. A quote should expire at the next table go-live, not on a round number. If the table flips at 00:00 Monday, the quote is good until 23:59 Sunday, and that time belongs on the quote in the same font as the price.

One change, two effects. You stop selling fuel you have not bought. And your sales team gets a reason to talk to the customer every week, which is a conversation they should be having anyway, because the customer's own buyer is getting asked the same question by their finance team.

Now the part that causes the arguments. Which timestamp decides which table applies. Pickup scan. Label creation. Manifest transmission. Handover at the counter. Every carrier's contract picks one, and the four can land on different sides of midnight for the same parcel.

Do not assume. Find the clause. It is usually one line in an attachment, and it is the line that decides whether a 22:40 pickup goes out at this week's number or next week's. Assume, for the sake of the example, that your agreement keys off the pickup scan and that your cutoff is 18:00. If the driver scans at 18:20, you are not twenty minutes late. You are a week of fuel late.

Watch the batch print as well. A warehouse that runs labels in one batch at 23:00 is printing four hundred labels on one side of a boundary or the other, and nobody in that building is thinking about a table. If your labels or your manifest carry a rate reference, the reference is whatever the system had loaded when the batch ran. Two hours earlier, different week, different number, same physical parcel.

Take the arithmetic, with the assumption stated out loud: base freight of 20 dollars a parcel, fuel charged as a percentage of that base. On the mainland tables that is 9.20 at 46.00, 9.35 at 46.75 and 8.70 at 43.50. On the DHL US-import table it is 7.35 at 36.75. On South America it is 6.60 at 33.00.

Same parcel. Same pickup. Same driver. The gap between the top mainland table and the South America table is 2.75 dollars a parcel, and geography created all of it.

Now put that in front of a picker at 17:00 with three orders and one van. The 2.75 does not change the routing decision, and it should not. What changes the decision is when the parcel has to be there. Fuel is not a routing variable at the counter, and pretending it is at 17:00 is how you miss the cutoff.

Seventeen hundred is the real decision point, not midnight. At 17:00 the route is set, the van is loaded, and the last three pickups are either on it or they are not. Nothing about a fuel table changes that decision, which is exactly why the cutoff has to be protected by a person rather than by a policy. A policy does not stand in a loading bay at 17:00 with a driver asking whether to wait.

Run the all-in number, because it changes how you talk. On the top mainland table, a 20 dollar base carries 9.35 of fuel. The invoice is 29.35, and fuel is 31.8 percent of what the customer pays. Nearly a third of the invoice is a line that resets every Monday.

If your account manager quotes 20 dollars and mentions fuel as an afterthought, the customer hears a 20 dollar shipment and receives a 29.35 one. That gap is not a pricing problem. It is a communication problem with a number attached, and it is why accounts go quiet in a week when nobody can explain the invoice.

Most mid-size shippers run one blended fuel percentage in the system. It is almost always wrong, and it is wrong in a direction you cannot predict. With tables running from 33.00 to 46.75 in the same week, the blend matches no parcel you actually ship.

At the counter this shows up as a margin report that says a lane is fine while the account manager says the lane is bleeding. Both are looking at the same week. One is reading the table, the other is reading the blend.

Here is the rhythm I would run, and it fits inside a week.

Monday at 00:00 the table flips. By 08:00, before the first pickup wave, someone has posted the five numbers in the ops channel with the week dates attached, not just the numbers on their own. By 09:30 every open quote in the system has been re-dated or re-priced. By 12:00 the accounts that ship weekly have a one line message with their new all-in per parcel, computed on their own lane mix rather than on the blend. By 16:00 the exceptions are closed and named: the fixed-price accounts, the rate that expires this month, the one that always argues.

That is four checkpoints and about ninety minutes of work. If it takes more than ninety minutes, the quote system is doing something wrong and you should fix that before you touch the fuel line.

The shift handover is where this dies. The person who prices sits on the early shift. The person who books sits on the late one. Between them there is a note, and the note says fuel updated, because someone typed it once at 09:00 and never looked again.

Say it in a form that survives a handover: table, week, top lane, cutoff. Five words. 46.75 on UPS, week of the 7th to the 13th, cutoff 18:00. Anything longer does not get read at 22:00. Anything shorter loses the cutoff, and the cutoff is the only part that cannot be recovered later.

Sooner or later a parcel gets held at a counter over a fuel dispute, and the argument is about nine dollars. The parcel matters more than nine dollars to your customer, and your customer's customer does not care about the dispute at all.

Get the parcel moving. Pay the difference, take the scan, close the manifest, and write down the timestamp and the number you paid. The argument can happen on Thursday with the invoice in hand. It cannot happen at 23:00 with a van waiting and a driver who has four more stops.

This is the one rule in here I would enforce without discussion. Speed first, reconcile second, and reconcile with a timestamp instead of a memory.

Keep three things when there is a dispute: the scan time, the table week you quoted against, and the published number for that week. Everything else is opinion. I have watched a nine dollar argument eat ninety minutes of a supervisor's evening because nobody could produce the second item, and the parcel sat there for all of it.

Carriers re-rate shipments after the fact. It happens. The invoice arrives with a fuel line that differs from the quote, small per parcel and large per week.

Two things make that survivable. Your quote carries the table week on it, so the dispute becomes a comparison of two documents instead of two recollections. And you keep a weekly file of the five numbers with the date in the filename, because the carrier's published table comes off the front page when the next week goes up, and after that you are arguing from a screenshot somebody took in a hurry.

By size, the work is different, and it is worth being honest about which size you are.

Say you ship 50 parcels a week. You do not need a system. You need one line in the quote template that carries the fuel number and its week, and you need to update it Monday morning before the first quote leaves the building. Fifteen minutes.

Say you ship 200 to 800 a week. You need the 09:30 checkpoint and one owner. One person whose name is on the re-dating, not a committee, because a committee re-dates nothing at 09:30 on a Monday.

Above that, you need the lane split, because a blend at that volume hides six figures a quarter, and you need the exception list written down, because at that volume there are always accounts on legacy pricing and they are always the ones that blow up on a Friday.

Split customers by the promise you made them, not by their size. Same-hour and next-flight promises cannot absorb a missed cutoff, so their all-in carries a buffer and nobody argues about it. Scheduled, consolidated promises can absorb 24 hours, and for those the cheapest answer is a later cutoff and a bigger batch. Same table, two different products, two different prices.

By lane mix, the split is between shippers who are all-mainland and shippers carrying a US-import or South America book. The all-mainland shipper pays the top of the range every week and has no hedge except service choice. The shipper with a mixed book has a real spread inside a single invoice, and most of them do not know it, because the blend flattened it.

There is a version of this that costs money quietly. A customer moves volume off a 46.75 lane onto a 36.75 lane for service reasons, and the fuel line drops by ten points without anybody noticing. That is not a saving you negotiated. It is a saving that happened to you, and it will disappear the same way it arrived.

The 33.00 and 36.75 tables are not a discount you can go shopping for. They belong to specific destination books, and you cannot move a parcel onto one by asking nicely at the counter. What you can do is check where your volume already sits, and stop quoting the top table across the whole book. If a third of your parcels are on a lower table, a blended quote at the top overprices that third and underprices nobody.

Every published summary says the same thing: above 40 for months, largest variable cost after base freight. True, and not the useful part.

The useful part is the reset frequency. A cost that is 46 percent and fixed for a year is a budget line. A cost that is 46 percent and resets every Monday is a process. You cannot budget a process with a number. You budget it with a rule, and the rule has to be executable by somebody at 17:00 on a Friday, which is the only test that matters.

That is what the tables do not tell you, and it is why two shippers with identical volume and identical lanes end up with completely different fuel outcomes. One has a rule. The other has a spreadsheet that gets updated when someone remembers.

Above 40 for months also means your customer's expectation was formed somewhere else. They remember fuel as a small line, or they remember a number from a year when it was a small line, and a new quote reads to them as a price increase you invented on a Tuesday.

The answer is not an argument. It is the table, forwarded with the week dates on it. A customer who sees the carrier's own published number stops negotiating with you and starts planning around it. Two minutes, and it changes the tone of the call.

The percentage is limited by the market. The cutoff is limited by nothing. A table can move 13.75 points across lanes and you can still plan around it. A missed cutoff costs you 24 hours and a phone call to somebody who has already printed the delivery promise on their own paperwork.

So the operational priority is the reverse of the financial one. Finance wants to talk about the percentage. Ops should be protecting the cutoff, because the cutoff is the only number in this whole story that you fully control.

Concretely, the next 72 hours. Before the first pickup on Monday, the five numbers are posted with the week attached. By 09:30 the open quotes are re-dated. By 12:00 the weekly accounts have their one line and their own per-parcel number. By 16:00 the exception list is closed, named, and assigned.

By Wednesday the quote template carries the new validity rule, so next Monday is not a rerun of this one. By Friday the five numbers are filed with the date in the filename, and the lane mix has been run once so you know what share of volume sits on the top table.

None of that is a project. It is one person, one hour a week, and permission to refuse a quote that has no table week on it.

A note on who owns this, because ownership is the part that quietly fails. In most shippers nobody owns the fuel line. It sits inside freight, gets looked at when somebody complains, and gets forecast by nobody. Give it a name and put it in the weekly ops meeting next to on-time performance. The moment fuel sits beside on-time, people stop treating the cutoff as a separate problem, which it never was.

Run the other direction. If the tables ease next Monday, everything above still applies, with better numbers. The re-dating still happens. The cutoff still matters. The customer still needs the number with the week attached to it.

What changes is your incentive to be fast. When the table falls, a slow re-quote is a gift to your customer and a cost to you, and the accounts that noticed you were slow to pass on the increase will notice you were slow to pass on the decrease. Same process, opposite reputation.

What flips the read is not the table. It is the base on your own invoice. Fuel is a percentage of a base freight number, so a 46.75 sitting on a base that has moved is a different amount of money from the same 46.75 on the base you budgeted. Anyone telling you the fuel line is fine because the percentage held still is telling you they did not check the base.

Alternatives, priced in hours. Chasing a lower table by switching carrier costs an account setup, a test shipment and a re-papered rate card, and it buys you a table that resets exactly as fast. Splitting volume across two carriers costs you the cutoff priority you spent a year earning. Consolidating parcels into fewer, bigger shipments saves fuel per parcel and costs 24 hours on everything you held back.

The trap that catches most people is quoting all-in without the table week on it. The next one is telling the customer that fuel is high right now instead of telling them the number and the week. The first is a contract problem. The second is a trust problem, and it takes longer to fix.

There is a version of this that is worse than doing nothing: quoting the table without the week. A quote that says 46.75 with no date on it looks precise and behaves worse than a range, because the customer reads it as a promise and your ops team reads it as history. Precision without a timestamp is the most expensive kind of ambiguity in this business.

By the time you read this, the week of 7 to 13 September is history and the table has moved again. That is fine. The number was never the point.

The point is that a number with a switch-on time needs a process with a clock, and the clock has to be yours, because theirs will keep flipping on Monday whether or not anyone in your office has a reminder set.

Get the parcel out. Send the number with the week on it. Argue on Thursday, with the file open.

↑ Back to the key points

— By Nadia Pryce