The Freight You Never Charged For: How Ferrous Yards Quietly Lose Money on Transport

There is a particular kind of deal that feels great on the phone and looks terrible three weeks later. A mill buyer needs volume. You quote a number per tonne. He agrees fast, which should have been the first warning. You load it out, the material moves, the invoice goes out, and the month closes lower than it should have. Nobody can point to the leak, because on paper every load sold above cost.

The leak is usually freight, and in ferrous it is rarely small.

Ferrous is a haulage business with metal attached

The reason freight bites harder here than in non-ferrous is arithmetic, not bad luck. A tonne of copper-bearing material can carry several hundred thousand rupees of value. A tonne of heavy melting scrap carries a small fraction of that. The truck does not care. It costs roughly the same to move either one across the same 120 kilometres.

So freight as a share of the value in the vehicle can run from almost nothing on a non-ferrous load to a meaningful slice of gross margin on a ferrous load. In many ferrous yards, transport is one of the two or three largest cost lines after the material itself, and it is the one most often treated as a monthly overhead lump rather than a cost belonging to a specific load, a specific grade, and a specific customer.

That accounting habit is where the money hides. If freight lives in a single line at the bottom of the P&L, every grade looks profitable and the mix decisions get made blind.

Cost per tonne-km is the only unit that travels

Freight quotes arrive in inconsistent shapes: a lump sum for the trip, a rate per kilometre, a rate per tonne, a rate for the vehicle for the day. None of those compare cleanly to each other, and none of them tell you whether a specific deal earns its keep.

Convert everything to one number: cost per tonne of material delivered, built from a per-trip cost and an assumed payload.

A simple, illustrative build for an owned or contracted trailer running a 120 km one-way haul:

At a 25 tonne payload, that is ₹525 per tonne. Expressed the other way, it is roughly ₹4.4 per tonne-kilometre of loaded distance. Pick one convention, write it down, and use it for every lane. Mixing loaded-km and total-km in the same comparison is a common way to convince yourself a distant customer is closer than they are.

The point of the per-tonne-km number is that it travels. It lets you answer, in ten seconds, whether a mill 210 km away at a higher price is actually better than a mill 90 km away at a lower one.

Payload is the denominator you control

Notice what happens when only the payload changes. The trip cost is fixed. The material fills the truck differently.

That spread is larger than the gross margin on some ferrous grades. Which means load optimisation is not a housekeeping topic. It is a pricing topic.

The practical work sits in a few places:

Know which grades weigh out and which cube out. Dense material such as heavy melt or shredded feed fills the weight limit before it fills the body. Light gauge sheet, tin, and loose turnings fill the body first and leave tonnes on the table. Those two categories deserve different freight assumptions, and often different vehicles.

Process before you move, when the maths supports it. Baling, shearing or cutting to length costs money per tonne. If it lifts an average load from 18 to 24 tonnes on a long lane, compare the processing cost per tonne against the ₹200 or so per tonne of freight you just recovered. Sometimes it pays and sometimes it does not. The mistake is deciding by habit instead of by lane.

Mix loads deliberately. A layer of dense material under light material can bring a cubing-out load closer to its weight limit, subject to what the buyer will accept on a single ticket.

Track achieved payload, not permitted payload. Yards routinely plan at the vehicle's rated capacity and quietly average several tonnes below it because of loading practice, moisture, or a rushed dispatch at the end of the shift. If you have never compared planned tonnes against actual weighbridge tonnes per trip, that gap is the first place to look.

The empty leg you are already paying for

Every round trip in the table above includes a return journey carrying nothing. That is not a rounding error. It is roughly half the distance.

Backhaul is the cheapest freight available to a yard, because the fixed trip cost is already committed. Suppose the same vehicle returning from a mill picks up an inbound purchase from a supplier near that route, and the value of that leg is worth ₹5,000 against the trip. The outbound freight per tonne on a 25 tonne load drops from ₹525 to roughly ₹325.

That change is large enough to decide a deal. It is also invisible unless purchase pickups and sales dispatches are planned by the same person looking at the same map on the same day. In most yards they are not: buying and selling run as separate conversations, and the truck goes back empty because nobody was looking.

You will not fill every return leg. Filling some of them, on your busiest lanes, is usually the single highest-return freight improvement available without buying anything.

Ex-yard pricing and the trap in the middle

Here is how the deal from the opening paragraph goes wrong.

You quote ex-yard, meaning the buyer arranges collection. Clean and simple. Then the buyer asks you to arrange it "just this once", or the material is sold delivered because that is what the mill prefers, or a broker sits in the middle and the terms are vague. The freight ends up on your side without ever entering the price you agreed.

Or a subtler version: you quote delivered, using a freight assumption from a lane you ran six months ago, at a payload you rarely hit, with no allowance for detention at the mill gate. The quote is not wrong by much. It is wrong by ₹200 or ₹300 a tonne, repeatedly, on your highest-volume grade.

Three habits stop most of it:

• Every quote states the term explicitly. Ex-yard or delivered, in writing, on the document the customer sees.
• Every delivered quote carries a freight line built from a current lane rate and a realistic payload, not a memory.
• Detention and waiting time are costed. A truck sitting four hours at a mill gate is trip cost with no tonnes attached. If a particular buyer routinely holds your vehicles, that is a real cost of doing business with them and it belongs in their price.

Documentation deserves a mention here too. Movements need the correct tax documentation travelling with the vehicle, and in India e-way bill validity is linked to distance, so long lanes and long detentions can create genuine compliance exposure on top of the cost exposure. Treat delays as a paperwork risk, not only a scheduling annoyance.

Seeing freight where it actually belongs: margin per grade

The fix at the reporting level is to allocate freight to the load, then roll it up by grade and by customer, instead of leaving it in a monthly bucket.

Consider an illustrative yard with three ferrous grades. Use your own numbers; the shape is what matters.

Before freight, light sheet looks healthier than turnings. After freight, it is the weakest of the three, because it cubes out and drags the payload down on every trip. A yard reading only the "gross before freight" column would happily buy more light sheet and wonder why the month keeps disappointing.

The same allocation applied by customer answers a question most yards never ask directly: which buyers are we effectively subsidising with transport? Usually there are one or two, and usually they are long-standing relationships nobody has re-priced in years.

Once the allocation exists, a few reports become worth reading every week:

• Freight cost per tonne by lane, trending
• Average achieved payload by vehicle and by grade
• Percentage of return legs carrying anything
• Margin after freight by grade and by customer

Where a system earns its place

None of this requires software in principle. It requires the underlying facts: what the vehicle actually carried, how far it actually went, how long it actually waited, and what it cost. In practice those facts are what go missing.

This is the part Scraplytics is built to make routine. Weighbridge-integrated scale tickets capture the real net weight per trip rather than a planned figure. Dispatch and the driver app record the trip itself, with GPS and proof of delivery, so lane distances and turnaround times are observed rather than remembered. Inventory tracked by category and grade with weighted-average costing gives you the cost side of the margin line. Reports and analytics then let you put freight next to grade, lane and customer instead of in a single overhead row, and invoicing with GST, e-way bill and e-invoice keeps the documentation moving with the vehicle.

The software is not the insight. The insight is that in ferrous, the truck is part of the product, and anything you do not price you eventually pay for.

Start here this week

Pick your three highest-volume lanes. For each one, work out the true round-trip cost, the average payload you actually achieved last month from your weighbridge records rather than the rated capacity, and the resulting cost per tonne. Put that number beside the margin you thought you were making on each grade moving on that lane.

Most yards find at least one grade, one lane or one customer that has been running at a loss in plain sight. Finding it costs an afternoon. Not finding it costs every month.