Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- By 19 June 2026, tender every shipment already scheduled for 22 to 30 June before the 21 June cut-off: on a 1.2 million dollar annual book that is about 29,600 dollars of freight at the old level, roughly 1,746 dollars saved, and do not pull forward freight that was not going to ship anyway, because nine days of carrying cost on cargo worth twelve times the freight works out to about 1,751 dollars.
- By 15 July 2026, get your top ten lanes priced individually in writing rather than as an average, and set the escalation trigger at 1.5 times the 5.9% headline, so any lane quoted above 8.85% goes back for renegotiation or moves to another carrier in the same week.
- By 31 July 2026, consolidate 20% of monthly shipments, 40 of an assumed 200, into half the number of bills, targeting about 4,236 dollars a month or 50,800 dollars a year at the assumed 1.6-times merged rate, with dock dwell on merged orders held under 48 hours.
- By 31 August 2026, run two lanes on rail intermodal for four weeks each at 15% of the book, with the delivery promise rewritten to the assumed seven to nine days and an on-time floor of 92%, targeting roughly 30% off the cost of that slice.
- By 30 September 2026, rebuild the 2027 freight budget on a ten-month increase cadence instead of a twelve-month one: at 5.9% each, that is 18.8% price level at the two-year mark instead of 12.1%, about 79,000 dollars of annualised run-rate on a 1.2 million dollar book.
- Set the reversal trigger now: if the Logistics Managers Index puts capacity back above 50 and the transportation prices reading below 60, stop adding contract cover and move volume back to spot, because the case for the faster cadence rests on that 64.3-point gap, not on the carrier's announcement.
ArcBest's LTL unit ABF Freight announced a 5.9% average general rate increase effective June 22, breaking the industry convention of annual-only price adjustments. The mid-cycle increase comes just 10 months after its previous August GRI and reflects tightening LTL market conditions: the May 2026 Logistics Managers Index recorded transportation prices at a 10-year high of 96.0 while capacity remained in contraction at 31.7. ArcBest also raised its Q2 outlook for both asset-based and asset-light units, with Q1 contract renewals averaging 6.3% increases, the highest since Q3 2022. The move signals that carriers are leveraging constrained post-Yellow capacity to reset pricing power more frequently, with implications for shipper budgeting cycles.
The Analysis
Most of my working life has been spent on corridors rather than on companies. I know the gauge change at Alashankou, the truck queue at Khorgos in February, and the freighter that rolls out of Zhengzhou at two in the morning and is on the ground in Leipzig before the European offices open. So when a pricing notice lands in front of me, the first thing I look at is never the percentage. It is the shape of the network underneath it. ArcBest's LTL arm, ABF Freight, has put out an average 5.9% general rate increase effective June 22, and the whole story is sitting in that one word, average. An average across a network is not a number that anybody actually pays. It is a spread. Some lanes will take three points, some will take nine, and the ones that take nine are the lanes nobody fights for.
What makes this worth ten minutes of your morning is not the 5.9%. It is the calendar. The industry has lived on one general rate increase a year for as long as most of the people signing these contracts have been in the job. ABF's previous one was in August, and this one lands in June, ten months later. Ten is not twelve. If the carrier holds that cadence, you are not budgeting for one reset a year, you are budgeting for one every ten months, and across a two-year window that is three increases where your model has two. I have watched the same thing happen on the rail side of Eurasia. A corridor that re-prices whenever it feels like it is no longer a tariff. It is a spot market wearing a contract's clothes.
Then there is the reading underneath the decision, and here I trust the Logistics Managers Index more than I trust any carrier's filing. In May 2026 transportation prices came in at 96.0, the highest in the index's ten years, while capacity sat at 31.7, still below the 50 line that separates growth from contraction. Prices are being reported at near-record expansion while the ability to move the freight is still shrinking. Those two legs of the same index are 64.3 points apart. ArcBest also lifted its second-quarter outlook for both its asset-based and its asset-light units, and its first-quarter contract renewals averaged 6.3%, the highest since the third quarter of 2022. Put those two numbers next to each other and the ordering is the interesting part: 6.3% in renewals is above the 5.9% in the published increase. The number on the announcement is the floor of what shippers actually sign, not the ceiling.
Let me draw the network first, because that is where this money actually hides. An LTL system is a road system with a spine. A pallet gets picked up on a local route, it goes into an origin terminal, it is unloaded and re-sorted, it rides a linehaul leg overnight, it comes off the truck again somewhere in the middle, and then it goes out on another local route to a door. Every one of those touches is a dock, and every dock is a place where a network short of space starts charging you for the privilege of standing still. I have never run a US LTL terminal and I am not going to pretend I have. But I have stood in enough transfer sheds between Urumqi and Duisburg to recognise the sound a network makes when it is short of doors, and it sounds exactly like this one: prices rising while the freight itself is not growing.
The mean trick in a rate increase quoted as an average is that it tells you nothing about your own book. Assume your top ten lanes carry 70% of your LTL spend and the long tail carries the other 30%. Assume the dense lanes, the ones a carrier wants because the truck was going that way anyway, come in at the published 5.9%. Assume the tail lanes, the ones feeding thin terminals and long local routes, come in at 1.4 times that. That 1.4 is my assumption, not a carrier's number, and you should argue with it.
The weighted result is 0.7 times 5.9 plus 0.3 times 8.26, which is 6.6%. On a 1.2 million dollar annual LTL book, which is the anchor I will use for every figure here and which you should replace with your own, that is 79,200 dollars against the 70,800 dollars you get if you take the headline at face value. Eight thousand four hundred dollars a year, created entirely by where your freight happens to sit on the map.
Now the calendar, because this is the part that should change how you write a budget. Ten months between general rate increases instead of twelve means that if the cadence holds, the next one lands around 22 April 2027. Assume each increase is 5.9%. Two of them inside a twelve-month window compound to 1.059 times 1.059, or 12.15% above where you are today. Three of them across a twenty-four month window reach 1.059 cubed, which is 18.77%. A budget written on the old annual rhythm expects about 12.1% at the two-year mark. The ten-month rhythm puts you at 18.8%. That 6.6-point gap in price level is, on the same 1.2 million dollar book, roughly 79,000 dollars of annualised run-rate that nobody has put in the plan. That is the number I would carry into a finance meeting. The 5.9% is the headline; the cadence is the story.
Why they can do it is not mysterious. Capacity is still tight after Yellow left the market, and the May 2026 Logistics Managers Index says so in two numbers that pull in opposite directions: transportation prices at 96.0, a ten-year high for the index, and capacity at 31.7, still under the 50 line that separates growth from contraction. When prices are reported at near-record expansion while capacity is still shrinking, a carrier does not need to be brave to raise prices in June. It only needs to be early. ArcBest raised its second-quarter outlook for both the asset-based and the asset-light units on the back of this, and its first-quarter contract renewals averaged 6.3%, the highest since the third quarter of 2022. Read those two together and the ordering matters more than the levels: renewals at 6.3% are running hotter than the published 5.9%. What gets announced is the floor of what shippers are signing, not the ceiling.
Between now and 22 June there is a window. It is small but it is real, because LTL is billed on the day you tender the freight, not the day it delivers. Assume that 1.2 million dollar book runs evenly across the year, which is 3,288 dollars a day. Pull nine days of freight forward, everything already scheduled for 22 to 30 June, and tender it on or before 21 June, and you have about 29,600 dollars of freight billed at the old level. At 5.9% that is roughly 1,746 dollars you do not pay.
Now do the other half of the arithmetic, because this is where most people fool themselves. Assume the cargo in those shipments is worth twelve times the freight, which is my assumption, so the value moving is about 355,200 dollars. Assume carrying cost runs 20% a year, another of my assumptions, high for most people and low for a few. Nine days of carrying that is 355,200 times 20% times 9 divided by 365, about 1,751 dollars. You have saved 1,746 dollars and spent 1,751 dollars.
So the rule is not pull everything forward. The rule is pull forward only the freight that was going to move in that window anyway, the orders already picked, already packed, already promised. Anything you manufacture early to chase a freight saving is inventory wearing a logistics costume, and it will cost you more in carrying cost and in dock congestion than the increase ever would. I have made this exact mistake on the rail side, booking space three weeks early to beat a rate change and then paying for a warehouse full of goods nobody had ordered. Say it out loud: the saving is in the timing of the tender, not in the timing of the production.
The lever I would actually push is consolidation, because it attacks a part of LTL pricing that has nothing to do with what the carrier pays for. Take my assumed 200 shipments a month at an assumed 500 dollars each. Assume 40 of them travel in the same direction in the same week and can be merged into 20. Assume a merged shipment costs 1.6 times a single shipment rather than double, because LTL pricing carries a minimum charge and rewards a heavier, denser unit. Twenty merged shipments at 1.6 times is 32 shipment-equivalents against the 40 you were running, so you retire eight bills. At the post-increase 529.50 dollars, that is about 4,236 dollars a month, roughly 50,800 dollars a year. No negotiation required, no relationship damaged, just fewer bills moving more weight each.
What it costs you is discipline at the dock. Merging shipments means somebody has to hold an order for two or three days and somebody has to tell a customer that the delivery date moved. If your promise to the customer is next-day, consolidation will break it. If your promise is this week, it will not. I would rather run the numbers on the second kind of promise and leave the first kind alone. Watch the dwell time at your own shipping dock while you do it: if merged shipments start sitting more than 48 hours waiting for a partner order, you have turned a freight saving into a warehouse problem.
Then there is the lever from my own side of the house, which is to stop treating the road as the only road. Take a long lane, say 1,200 miles, and assume LTL runs it door to door in four to six days. Assume rail intermodal runs the same lane in seven to nine days. Assume intermodal costs 30% less on that lane. All three are my assumptions and you will need your own quotes, but the shape of the trade is stable: you buy two to four days with roughly a third of the cost. Move 15% of the book, 15,000 dollars a month on my anchor, and you are looking at about 4,500 dollars a month, some 54,000 dollars a year, paid for in days. On a lane where nobody is standing around waiting for the goods, that is the easiest money in this whole exercise.
It only works under conditions, and the conditions are physical. You need a ramp at both ends that your drayage carrier actually serves, you need enough weight to fill the box rather than pay to move air in it, and you need to accept that rail schedules breathe in a way truck schedules do not. Think of it as changing a tyre rather than changing the car. You keep the same cargo, the same customer, the same invoice, and you swap the middle of the journey. The mistake everybody makes is swapping the middle and leaving the promise alone. Rewrite the delivery promise on those lanes to the seven-to-nine-day range before you move a single pallet, and hold an on-time floor of 92% as the go or no-go for widening the trial.
Air, since somebody always raises it. Air is not an answer to a rate increase, it is an answer to a line going down. Assume air runs twelve times the cost of this road move per kilo, my assumption, generous for short lanes and brutal for long ones. Switching costs you eleven times the freight bill on top of what you were already paying. That trade only makes sense when the cost of being late is more than eleven times the cost of moving the box. For a pallet of machine parts holding up a production line, that can be true by Tuesday. For a pallet of anything else it is almost never true, and the people who reach for air in June are the people explaining a freight budget in September.
There is a fourth option that costs nothing and takes a week to arrange, which is to change your rhythm. Ship twice a week instead of every day, in fuller loads, on fixed days. What you are doing is not saving money on the rate, you are changing the profile of what you tender: fewer bills, a heavier average weight per bill, a predictable pick-up a carrier can build a route around. Carriers have always priced predictability cheaper than they price surprises. Assume you move from 200 bills a month to 120 bills carrying the same weight, and every pricing mechanism in LTL, from the minimum charge upward, starts working in your favour rather than against you.
A network under stress behaves like a ferry crossing on a holiday weekend, and I lean on that picture because it tells you where to stand. The ferry does not get faster because you are in a hurry. It takes the cars already in the queue, in the order they arrived, and the cost of the crossing is set by how many people want to get on. LTL in a tight market is the same: what you pay is decided less by your contract than by whether your freight is tendered early enough in the day to make that night's linehaul.
Miss the cut-off by an hour and the pallet spends a night on the dock, and a pallet that spends Friday night on the dock is now competing for Monday's space at Monday's price. Go and get the actual cut-off times at every terminal you use, written down, this week. Most shippers I sit with cannot produce them, which means they are negotiating rates against a schedule they have never read.
One more thing about that 6.3%, because it is the most usable number in the whole story. It is a first-quarter number, and it sits above the increase that was only announced in June. That means the shippers who renewed in the first quarter were signing above the public figure months before the carrier put anything on the record. If your renewal falls in the first quarter, you are in that group. The fix is boring but it works: stagger the renewals. Instead of every contract expiring in the same quarter, split the book so no more than a third of it reprices in any one quarter.
Assume you have three carrier relationships and all three come up in March; move one to May and one to September, even if it costs you a short extension to get there. What you give up is the chance to play all three against each other in a single week. What you get is that no single quarter can put a 6.3% average across your entire book.
The traps are mostly traps of assumption. The one I see most is believing the average is your number, and it is not. The second is assuming a negotiated contract lands under the published increase; first-quarter renewals at 6.3% against a 5.9% announcement say the opposite, and if your renewal falls in the first quarter you are already inside that statistic. The third is billing weight and the minimum charge, which is where a consolidation programme quietly fails: you merge two shipments, the density improves, and a minimum charge eats the difference, so ask for the merged shipments to be rated rather than estimated. The fourth is the one that costs the most, assuming the next increase is a year away and writing a flat line into next year's plan. Ten months is not a year, and the difference compounds.
On who does what, I would keep it tight and put a name next to a date. The transportation manager owns the tender calendar and the lane-by-lane pricing request, and I want those lane quotes in hand by 15 July, not the average, with anything above 8.85% flagged for renegotiation the same week. The warehouse owns consolidation and the dock dwell metric, with a target of 20% of bills merged by 31 July. Procurement owns the 2027 budget rework, due 30 September, built on a ten-month cadence instead of a twelve-month one. Customer service owns the promise rewrite on any lane moving to intermodal, before the first pallet moves, not after the first complaint.
Now, what would make me wrong. The whole case for a ten-month cadence rests on the gap between those two index readings, 64.3 points between prices and capacity. If capacity climbs back over 50 and the transportation prices reading falls below 60, the pressure comes off and carriers start buying volume instead of pricing it. Capacity at 31.7 is deep in contraction and will not turn in a month, but it does move. The other thing that would turn this is weight per shipment. Carriers price off weight as much as off count, and a book that ships the same number of bills at a lower average weight is a book that gets re-rated. Track your own weight per bill monthly. If it is drifting down, you are the reason your own rates are going up.
And tell your own people before 22 June, not after. The failure I have watched a dozen times is not a bad rate, it is a sales team still quoting June numbers in August because nobody told them the tariff moved. Assume freight runs 8% of your cost of goods sold, my assumption, and assume your quotes carry a 60-day validity; every quote issued after 22 June on the old numbers is a margin leak you will find a quarter later, on the invoices, when it is too late to reprice the order. Set a date, 19 June, on which the transport team hands commercial a single page with the new lane rates and the day they start, and make it a standing habit rather than a one-off exercise. The carriers have turned this into a twice-a-year event. Your own internal notice ought to be at least as reliable as theirs.
Ten days to the effective date, one number that is not your number, and a cadence that no longer matches the calendar in your budget. That is the whole of it. I have worked corridors where the tariff moved twice in a season and corridors where nobody dared move it for three years, and the difference was never the carrier's nerve. It was whether the road had room. Prices at 96.0 and capacity at 31.7 is a road without room. Say it plainly: this is a road problem, and road problems get solved with routes, loads and calendars, not with arguing about percentages.
↑ Back to the key points