Daniel's been staring at an AliExpress tracking page again. He wants to know why the package tears across the Pacific in two days, clears customs, reaches a distribution center, and then just... sits there. The last two hops. And he's asking what line haul actually is, how LTL fits into the final stages, and what happens when you scale this to a country the size of the U.S., where there's no single Tel Aviv to Jerusalem route but thousands of them.
And he's right that the slowdown is real. The international freight leg is optimized to the point of being boring. The destination side is where everything fragments.
So let's trace a package from the port to the doorstep and see where the system gets weird.
Last-mile logistics is the final leg, from a distribution hub to the end customer. And the thing most people don't realize is it's the most expensive part of the entire supply chain. Depending on whose numbers you use, the last mile is somewhere between forty and fifty-five percent of total shipping cost.
More than the ocean crossing.
Much more. The ocean crossing is a container with eight thousand identical boxes stacked forty feet high. The last mile is a person in a van making ninety stops, most of which involve finding parking, buzzing an intercom, and waiting for someone who isn't home.
And that's where LTL becomes relevant. Less-than-truckload. You're not filling a truck with one shipper's goods. You're consolidating parcels from dozens or hundreds of shippers into shared truck space. The final mile is almost never a full truckload, so LTL is the default model.
The core tension is this. The international freight leg is built for scale. Massive ships, massive containers, massive consolidation. Then the package hits the destination country and has to be broken down into smaller and smaller units. Each breakdown adds a handoff. Each handoff adds time.
And a tracking status that makes you think the package vanished.
Let's walk through the actual stages. Port to national distribution center. National distribution center to regional hub. Regional hub to line haul between cities. Line haul to local delivery or pickup point. Each of those transitions is a sorting event.
When you look at an AliExpress tracking page, the status that catches people is "Arrived at Linehaul Office." That's the marker where the package transitions from long-haul to the last two hops. And it's where packages sit for days.
Because the line haul office is a consolidation point. Think about what has to happen for a truck to go from Tel Aviv to Jerusalem. The carrier doesn't want to send a truck with twelve parcels. They want a truck with eight hundred parcels. So the line haul office accumulates packages until there's enough volume to justify the run.
So the package isn't moving. It's waiting for friends.
And that waiting period is invisible to the customer because the tracking just shows a static status. "Arrived at Linehaul Office." It sounds like a place. It's really a queue.
Then the second hop is even slower. The package arrives at the Jerusalem facility, gets re-sorted by neighborhood, and handed to a local courier. That courier is one driver navigating traffic, making stops, dealing with failed delivery attempts. A single driver might handle sixty or eighty packages a day. If you're number seventy-nine, you wait.
The failed delivery attempt is the hidden cost multiplier. In some markets, failed first-attempt delivery rates run fifteen to twenty percent. Every failure means the driver comes back the next day. That's not a logistics problem, that's a scheduling problem.
So let's get into the mechanics of LTL at the destination side. A distribution center aggregates parcels from multiple shippers. Sorts them by destination city. Loads them onto line-haul trucks. At the destination city, parcels are unloaded, re-sorted by neighborhood, and handed to local couriers.
Every one of those steps is a handoff. And every handoff is a potential delay. The parcel gets scanned in, scanned out, moved to a new staging area, loaded onto a different vehicle. If any one of those scans is missed, the tracking looks broken even though the package is moving.
The "Arrived at Linehaul Office" status is the one Daniel noticed because it's the most visible bottleneck. The system is deliberately holding packages to consolidate enough volume for an efficient run. The tradeoff is speed versus cost.
And the economics are brutal. A line-haul truck from Tel Aviv to Jerusalem costs roughly the same whether it's full or half empty. Fuel, driver, maintenance, tolls. So the carrier has every incentive to wait until the truck is full. The customer has every incentive to want it to leave now. Those incentives never align.
What's the actual cost difference? If a carrier sends a truck at forty percent capacity versus ninety percent?
I don't have exact figures for Israeli carriers, but the general rule in trucking is that a line-haul truck needs to run at about seventy to eighty percent capacity to be profitable on that leg. Below that, the carrier is losing money on the route. So the wait isn't arbitrary. It's the carrier protecting the margin.
And the customer is paying for shipping that was priced assuming a full truck.
Right. The AliExpress model works because the shipping is either free or nearly free, which means the carrier has to squeeze every efficiency out of the line haul. They can't afford to send a quarter-full truck from Tel Aviv to Jerusalem. So the package waits.
Let's talk about the sorting process itself. At the line haul office, what's actually happening to the parcel?
It arrives in a bin or a cage with other parcels from the same origin facility. Workers scan it, verify the destination, and place it in a staging area for the next truck. In a well-run facility, that takes hours. In a busy facility, it can take a day or more. Then it sits in the staging area until the truck is loaded.
And the truck might not leave for another day or two while volume accumulates.
So the "last two hops" Daniel is seeing are really three sub-stages. Arrival at the line haul office, waiting for consolidation, then the actual line haul run. The tracking status collapses all three into one line.
That's the misconception. Tracking updates make you think the package is moving. "Arrived at Linehaul Office" sounds like progress. It can mean the package is sitting on a pallet in a warehouse for forty-eight hours.
And the final mile has its own delays. The local courier receives the package, but the delivery route is optimized for the courier's efficiency, not the customer's urgency. If you're at the end of the route, you wait.
What about pickup points? Daniel mentioned them. How do they change the equation?
Pickup points are the workaround. Instead of the courier making ninety individual stops, the courier drops thirty packages at a pickup point. The pickup point holds them until customers come. That eliminates the failed delivery attempt problem entirely.
The customer does the final mile themselves.
And the customer doesn't mind because they can pick up the package when it's convenient. The pickup point is open longer hours than a home delivery window. In Israel, pickup points are everywhere now. Supermarkets, convenience stores, kiosks.
It's not an upgrade from home delivery. It's an upgrade from never receiving packages at all.
Right. If you're not home during the day, home delivery means a failed attempt and a trip to the post office anyway. The pickup point just cuts out the failed attempt.
So the two-hop model in Israel is distribution center to line haul, line haul to local courier or pickup point. Two hops. Now scale it to the U.S.
The U.S. is where this model explodes in complexity. Instead of one main intercity route, there are thousands. Los Angeles to Phoenix. Chicago to St. Louis. Dallas to Houston. Every one of those is a line haul leg with its own consolidation logic.
A package from the Port of Long Beach to a rural town in Ohio might pass through four or five hubs.
Let's trace it. The container arrives at Long Beach. A drayage carrier moves it to a Los Angeles area distribution hub. That's hop one. The package is sorted and loaded onto a line-haul truck to Chicago. That's hop two. In Chicago, it's unloaded, re-sorted, and loaded onto another truck to Columbus. Hop three. In Columbus, it's re-sorted again and handed to a regional carrier. Hop four. The regional carrier takes it to a local post office or courier. Hop five.
Five hops, five handoffs, potentially five different carriers.
Each with its own tracking system. The port drayage carrier has one system. The national line-haul carrier has another. The regional LTL carrier has a third. The local courier has a fourth. When the package changes hands, the tracking often goes dark for a day because the new carrier's system hasn't updated yet.
So in Israel, two hops and maybe two carriers. In the U.S., five hops and four carriers. That's why U.S. delivery times are less predictable.
And why the consolidation problem is harder. In Israel, the Tel Aviv to Jerusalem line haul is a single route with predictable volume. In the U.S., the Los Angeles to Chicago line haul is competing with dozens of other origin-destination pairs. The carrier has to balance volume across a network, not just fill one truck.
That's the knock-on effect. The consolidation logic doesn't just scale linearly. It scales with the number of routes, which grows roughly with the square of the number of hubs.
Right. If you have ten hubs, you have forty-five possible routes between them. If you have fifty hubs, you have over twelve hundred possible routes. Each route needs enough volume to justify a truck. That's a much harder optimization problem.
And the customer sees the result. A package that moves quickly through the international leg and then stalls at a regional hub in the middle of the country.
The tracking fragmentation is real. I've had packages where the tracking showed "In Transit" for five days with no update, and then suddenly it was out for delivery three states away. What happened was the package moved through two carrier handoffs and neither system synced properly.
More carriers don't mean faster delivery. They mean more points of failure.
That's the misconception. People think competition between carriers speeds things up. In practice, each handoff is a place where the package can sit, get misrouted, or lose its tracking thread.
Let's talk about pickup points and lockers in the U.S. context. How are they changing the final mile?
Amazon Locker is the big one. Amazon has thousands of locker locations in convenience stores, gas stations, apartment buildings. The driver drops twenty packages at one locker bank instead of making twenty stops. That's a massive efficiency gain.
UPS Access Point is the same idea. Drop packages at a local business that's already open, let customers pick them up.
The economics are compelling. A failed home delivery costs the carrier something like fifteen to twenty dollars when you factor in the return trip and the re-delivery attempt. A locker delivery costs a fraction of that because it's one stop with a guaranteed successful drop.
So the locker network is essentially turning the final mile into a mini line haul. Consolidating many deliveries into one stop.
That's the insight. The final mile is becoming more like the line haul. Instead of individual stops, you have consolidated drop points. The customer becomes the last leg.
Which raises the question of whether the model is converging. If pickup points and lockers keep growing, the final mile might eventually look like a two-hop system even in the U.S. Distribution hub to locker. Customer does the rest.
But there's a counter-pressure. Same-day delivery. Amazon, Walmart, Target are all pushing toward faster delivery, which fragments the final mile again. You can't consolidate same-day deliveries into a locker run that happens once a day. You need multiple runs.
So the push for speed works against the push for efficiency.
Same-day delivery means the truck leaves partially full because the customer won't wait for consolidation. That's the opposite of LTL logic.
Let's go back to the AliExpress case for a minute. Daniel's observation is that the outward consolidation and international freight are fast, and the last two hops are slow. Why is the international leg so much faster?
Because it's been optimized for decades. The container shipping industry has standardized everything. Container sizes, port operations, documentation. A container moves from Shenzhen to Tel Aviv in a predictable window because the whole system is built for scale.
And the destination side is inherently fragmented. Every city, every neighborhood, every building has its own delivery constraints.
The international leg is a conveyor belt. The last mile is a maze. You can't optimize a maze the way you optimize a conveyor belt.
What's the actual time breakdown for a typical AliExpress order to Israel?
The international freight leg, from consolidation in Shenzhen to arrival at the Israeli distribution center, might be five to seven days. Then the last two hops add another three to five days. So the last mile is roughly half the total delivery time.
And that's in a country the size of New Jersey.
Israel is about the same area as New Jersey, yes. So the two-hop model works because the geography is small. Tel Aviv to Jerusalem is about sixty kilometers. That's a morning drive.
In the U.S., the equivalent line haul might be Los Angeles to Phoenix. Six hundred kilometers. Or Chicago to St. Louis. Five hundred kilometers. The distances alone add a day per hop.
And the U.S. has far more population centers. Israel has a handful of major cities. The U.S. has dozens. Each one is a potential hub in the network.
Let's talk about the carrier structure in the U.S. LTL market. Who are the players?
The big national LTL carriers are names like FedEx Freight, Old Dominion, XPO, Estes, Saia. They run networks of regional terminals. A package moves from terminal to terminal across the country, getting consolidated and deconsolidated at each stop.
And then the final mile is often handed off to a different entity entirely. The postal service, a local courier, an Amazon delivery partner.
The postal service is the default final-mile carrier for a lot of LTL freight because they already go to every address every day. The marginal cost of adding one more package to a postal route is tiny.
So the U.S. model is LTL line haul plus postal final mile. Two different systems stitched together.
And the stitching is where the tracking falls apart. The LTL carrier's tracking ends when the package is handed to the postal service. The postal service has its own tracking. The two don't always talk to each other.
That's the "tracking went dark" phenomenon.
Right. The package didn't vanish. It changed hands. But the customer has no way to see that.
What about the cost structure? How does LTL pricing work in the U.S.?
LTL pricing is based on freight class, weight, and distance. The freight class is a density-based classification. Dense, heavy items ship cheaper per pound than light, bulky items. A box of books is class sixty. A box of pillows might be class two hundred fifty.
So the same weight ships at very different rates depending on what it is.
And the pricing is per hundred pounds, per hundredweight. So a five hundred pound shipment at class sixty going five hundred miles might cost two hundred dollars. The same weight at class two fifty might cost five hundred.
That's the hidden complexity of LTL. It's not just sharing truck space. It's a pricing system that accounts for how efficiently the freight uses that space.
And in the last mile, that pricing gets absorbed into the delivery fee. The customer doesn't see the freight class. They see a shipping charge or free shipping. But the carrier is doing the freight class math behind the scenes.
Let's bring this back to Daniel's question about the U.S. scaling. He asked how the layers between port and doorstep work when there are countless intercity routes.
The key insight is that the U.S. doesn't have a single backbone. It has a mesh. A package from Long Beach to rural Ohio doesn't follow one route. It follows a path through the network, and that path depends on volume, capacity, and carrier contracts at any given moment.
So the same origin and destination might route through different hubs on different days.
Right. If the Chicago hub is congested, the carrier might route through Indianapolis instead. If the Columbus terminal is short-staffed, the package might go to Cincinnati and backtrack. The path isn't predetermined.
That's why U.S. tracking is so unpredictable. The customer sees a package in Chicago and assumes it's going straight to Ohio. Then it shows up in Indianapolis and the customer thinks it's lost.
It's not lost. It's being routed around a bottleneck. The network is making a judgment call.
And each of those judgment calls adds a day.
Each one adds a day and a handoff. The package might touch five or six facilities before it reaches the final mile carrier.
Let's talk about the pickup point trend in the U.S. more specifically. Amazon Locker, UPS Access Point, FedEx OnSite. How big is this?
Amazon has over nine hundred thousand delivery drivers and a massive locker network. UPS Access Point has tens of thousands of locations in the U.S., mostly in convenience stores and pharmacies. FedEx OnSite is similar. The trend is toward consolidating the final mile into fewer, denser stops.
And it's working. Failed delivery rates drop when packages go to lockers.
The failed delivery rate for home delivery is somewhere in the mid-teens percentage-wise. For locker delivery, it's close to zero. The package is either in the locker or it's not. There's no "customer wasn't home."
So the locker network is solving the most expensive problem in the last mile, which is the failed attempt.
And it's changing the economics. A locker delivery might cost the carrier a dollar or two in locker fees. A failed home delivery costs fifteen to twenty dollars in re-delivery. The math is obvious.
Which is why carriers are pushing pickup points so aggressively. It's not a customer convenience feature. It's a cost reduction strategy.
Both. The customer gets a more reliable delivery. The carrier gets a cheaper delivery. The pickup point operator gets foot traffic. Everyone wins except the person who really wanted it at their door.
Let's contrast the two models directly. Israel: one line haul, one final mile, two hops, maybe two carriers. U.S.: multiple line hauls, multiple regional hubs, multiple carriers, five or six hops.
The U.S. model is inherently less predictable. More hops means more variance. Each hop has a distribution of possible delays. The variances add up.
A package that could arrive in three days might arrive in seven, and the difference is just which hubs it happened to route through.
The customer experiences that as random delivery times. The carrier experiences it as network optimization.
What's the actual delivery time range for a Long Beach to rural Ohio package?
If everything goes perfectly, maybe three to four days. If there's congestion at a hub, a missed handoff, a weather delay, it could be seven to ten days. The variance is huge.
The tracking will show a burst of activity, then silence, then a burst, then silence. Each silence is a handoff.
The silence is the package sitting in a staging area waiting for the next leg. It's not lost. It's queued.
That's the thing I want listeners to take away. A static tracking status doesn't mean the package is stuck. It means the package is waiting for the next consolidation event.
The system is optimized for efficiency, not for visibility. The customer wants to see progress. The carrier wants to see full trucks.
Let's talk about what happens at the destination city. The package arrives at the regional hub, gets unloaded, re-sorted by neighborhood. What does that sorting look like?
In a modern facility, it's automated. Conveyor belts, barcode scanners, sorting machines that divert packages to different chutes. Each chute corresponds to a delivery route. The package gets scanned, sent down the right chute, and lands in a bin for the courier who handles that route.
That's where the neighborhood-level routing happens.
Right. The courier's route is designed to minimize driving distance and left turns. The package is assigned to a route based on its address. The courier loads the van in reverse delivery order so the first stop is at the back.
The amount of optimization in that final load is impressive.
It's a traveling salesman problem solved every morning for every route. The software has gotten very good at it. But the physical constraints remain. Traffic, parking, building access, customers not answering buzzers.
Even with perfect sorting, the final mile is slow because the physical world is slow.
The final mile is slow because it's the only part of the journey that happens one package at a time. Everything before that is bulk. The final mile is individual.
Let's pivot to the open question. As pickup points and locker networks grow, will the final mile become more like the line haul, consolidated and efficient? Or will same-day delivery fragment it further?
I think both are happening simultaneously. The standard delivery is getting more consolidated through lockers and pickup points. The premium delivery is getting more fragmented through same-day and on-demand services.
The market is bifurcating. Standard delivery gets cheaper and more efficient. Premium delivery gets faster and more expensive.
The customer chooses which one they want for each order. Free shipping to a locker, or pay extra for same-day to the door.
That's a clean split. The question is whether the same-day model can survive without massive subsidy.
Same-day delivery is expensive. The carrier can't consolidate, so the cost per package is high. Amazon absorbs some of that cost as a customer acquisition expense. But it's not sustainable as a standalone business.
The autonomous delivery angle is interesting. If drones or sidewalk robots handle the final mile, the driver bottleneck disappears. But the consolidation problem at the line-haul level remains.
Right. A drone can deliver one package from a hub to a doorstep in minutes. But getting the package to the hub still requires the line haul. The consolidation logic doesn't change.
The autonomous revolution, if it comes, changes the last hundred meters, not the last hundred kilometers.
The line haul is already efficient. The final mile is where the cost and complexity live. Autonomous delivery attacks the final mile directly.
But the line-haul consolidation bottleneck, the "Arrived at Linehaul Office" wait, that stays.
Unless the volume becomes so predictable that the truck leaves on a schedule regardless of fill level. But that requires the carrier to accept lower margins on some runs.
Which they won't do voluntarily.
Not unless competition forces it. And in most markets, the LTL carriers have enough pricing power to protect their margins.
Let's wrap the mechanics before we move on. The last mile isn't just the final drive to the doorstep. It's line haul, local sortation, and final delivery. Each stage has its own delays. The tracking statuses collapse all of that into a few lines.
The "Arrived at Linehaul Office" status is the one that confuses people because it sounds like progress but often means waiting. The package is in a queue, not on a truck.
Next time your package sits at a line-haul office, you'll know it's just waiting for a friend to share the ride.
That's the whole LTL model in one line.
I used to dispatch for a regional LTL carrier out of Indianapolis. Late nineties. Ran line-haul between Indy and Columbus. The consolidation bottleneck you're describing is real, but there's a human layer you're missing.
What's the human layer?
The dispatcher. The guy deciding which packages wait and which ones go. It's not just volume math. It's judgment calls. You've got a truck leaving at nine PM with sixty percent capacity. Do you hold the other forty percent for tomorrow and risk the customer calling, or do you send the truck and eat the margin? Every night, that's a person making that call.
The wait isn't purely algorithmic.
It's a guy with a clipboard and a phone, looking at a board of destinations and trying to guess what's coming in overnight. Sometimes he guesses wrong. Sometimes he sends a truck half empty because a big customer is screaming. Sometimes he holds packages because he thinks a big drop is coming and it never shows up.
What was the worst guess you ever saw?
We had a sorting facility in a converted bowling alley. The lanes were still there. The sorting bins were arranged along the old lanes, and they were labeled with city names in alphabetical order. So Columbus and Cleveland were right next to each other. Cleveland, Ohio, and Columbus, Ohio. Different ends of the state, adjacent bins.
That's a recipe for misrouting.
Happened constantly. A package for Columbus would fall into the Cleveland bin because the guy throwing packages was moving fast and the labels were two inches apart. We sent a shipment of live chicks on a two hundred mile detour once. Nineteen ninety-four. They got to Columbus three days late. The customer was not happy.
Live chicks.
They survived. Most of them. But the incident became company legend. The owner put a sign on the Cleveland bin that said "Not Columbus." Didn't help.
The bins were optimized for alphabetical lookup, not for package flow.
Right. Someone in management thought it would be easier to find the right bin if they were in alphabetical order. But the people actually throwing packages were working from route lists, not looking things up. They needed geographic grouping. Cleveland and Columbus should have been on opposite ends of the building.
That's a perfect metaphor for the last mile. The system is often optimized for internal convenience, not for package flow.
That's what I'm saying. You look at the tracking status and think there's some grand algorithm running. There is, at the big carriers. But at the regional level, it's a lot of tired people making calls at midnight in a building that used to be a bowling alley.
The bowling alley detail is the part I can't get past. Were the lanes still oiled?
No, they'd ripped out the wood. But the floor had these long grooves where the lanes used to be. Forklifts kept getting stuck in them.
That's absurd.
It was the cheapest lease in Indianapolis. That's why we were there. The company was running on thin margins, like every regional LTL carrier. You take what you can get.
The human layer changes how I think about the consolidation wait. It's not just waiting for volume. It's waiting for a dispatcher to decide the volume is sufficient.
The dispatcher is under pressure from both sides. The customer wants the package now. The owner wants the truck full. The driver wants to leave on time so he can get home. Three different incentives, one person in the middle.
That's the part the tracking page never shows.
The tracking page shows a status. "Arrived at Linehaul Office." It doesn't show the dispatcher looking at a half-empty trailer and deciding whether to wait for the midnight drop from the Chicago run.
The midnight drop might be late. Or it might not come at all. So the dispatcher holds packages for a truck that never fills.
Then the next morning, the packages go out on a truck that's still half empty, and the whole cycle starts again. That's why some packages sit for three days at a line-haul office. It's not one decision. It's a series of decisions, each one waiting for the next piece of information.
The system is less deterministic than people think.
Much less. The big carriers have software that predicts volume and optimizes routes. But the regional carriers, the ones handling the last-mile handoffs in a lot of places, they're still running on experience and guesswork.
Does that explain some of the tracking unpredictability Daniel sees?
Partly. The international leg is highly deterministic. Container ships run on schedules. Ports have clearances. The destination side is messier because it's local. Local volume, local traffic, local judgment calls.
The bowling alley is the perfect symbol for that messiness. A system designed for one thing, repurposed for another, with the old structure still causing problems.
Nobody ever fixed the bin labels. The "Not Columbus" sign was still there when I left.
That's the most human detail of all. The workaround becomes permanent.
The workaround is always permanent. That's the rule.
The open question I'm left with is whether the pickup point and locker trend will eventually remove the dispatcher's judgment call from the equation. If volume becomes predictable enough, the truck leaves on a schedule and the consolidation wait disappears.
It won't disappear. The volume is never perfectly predictable. But it could shrink. The locker networks give carriers a buffer. If the truck is half empty, the packages still have somewhere to go that's cheap to deliver.
The locker is the safety valve for the dispatcher's bad guesses.
The dispatcher can send the truck half empty because the locker delivery doesn't cost much. The margin hit is smaller. That changes the calculus.
That's a interesting interaction. The pickup point model doesn't just reduce failed deliveries. It reduces the cost of sending a partially full line-haul truck.
Because the final mile is cheaper, the line haul can afford to be less efficient. The consolidation pressure relaxes.
Which means the "Arrived at Linehaul Office" wait might actually shrink as lockers proliferate. Not because the line haul gets better, but because the penalty for sending a partially full truck gets smaller.
That's the knock-on effect. The locker network changes the economics of the entire chain, not just the final stop.
That's a good place to land. The final mile is becoming more like the line haul, and the line haul is becoming more forgiving. The whole system is converging toward consolidation at every level.
The customer sees fewer days of silence on the tracking page.
Fewer days of silence, fewer failed attempts, fewer trips to the post office. The pickup point is quietly fixing the last mile from the customer's end.
Next time your package sits at a line-haul office, remember the dispatcher in the bowling alley, staring at a half-empty trailer, deciding whether to wait for the midnight drop.
Be grateful the chicks eventually made it to Columbus.
Most of them did.
This has been My Weird Prompts. Thanks to Hilbert Flumingtop for producing, and for the bowling alley dispatch story.
If you enjoyed this, email us at show at my weird prompts dot com. Or visit my weird prompts dot com for more episodes.
We'll be back soon.
