# Short Shipments: Why the Carton Count Never Matches the Cut Quantity

> The arithmetic of cut to sewn to passed to packed to dispatched, where the leakage happens at each step, how a CUT / RECEIVED / DISPATCHED lot ledger works, and what to check when a buyer reports a shortfall.

**Source:** [https://scanerp.pro/blog/dispatch-accuracy-short-shipment-garment.html](https://scanerp.pro/blog/dispatch-accuracy-short-shipment-garment.html)

---

# Short Shipments: Why the Carton Count Never Matches the Cut Quantity

Santosh Rijal · September 16, 2026 · 9 min read · Dispatch

**TL;DR — Direct Answer:**
The cut quantity and the shipped quantity are never equal, and the gap is not one leak but five: **cutting issue, sewing loss, checking rejection, packing loss and dispatch error**. Most factories only measure the two ends, which tells you the size of the problem and nothing about its location. The fix is a **lot pipeline ledger** — running CUT, RECEIVED and DISPATCHED balances per lot and article — so the two gaps are separated: cut minus received is a production loss, received minus dispatched is a finishing and packing loss. **B-grade and rework distort the count** because graded-out pieces reduce the shippable quantity without being scrap, and pieces in the repair queue are counted as made but are not available to pack.

A buyer emails to say the shipment is 153 pieces short against the purchase order. The cutting sheet says 5,000 were cut. Dispatch says 4,847 went into the container. Both numbers are correct. Nobody in the building can explain the 153, and the conversation immediately becomes an argument between departments rather than an investigation.

I run a CMT factory in Nepal and I have had that email more times than I would like. This article works through where the pieces actually go, why the two-number comparison cannot answer the question, and the ledger structure that can.

## The Arithmetic Nobody Writes Down

The pipeline from cutting table to container has five stages, and each one has a characteristic way of losing pieces.

### Cut to issued

The cutting sheet says a quantity; what physically leaves the cutting room in bundles is sometimes less. Panels get rejected at the cutting table for a fabric fault caught during spreading, a layer comes out short because the marker did not fit as planned, a bundle is made up with a miscount and the ticket says forty when the bundle holds thirty-eight.

This loss is the most likely to be invisible, because in a lot of factories the cut quantity on the sheet is treated as fact and never independently verified. If the sheet says 5,000 and the room issued 4,972, every downstream number is being compared against a figure that was never true. That is why the first thing to check in a shortfall investigation is not the packing floor but the cutting record. We have written about the upstream side of this in the guide to [cutting room management](/blog/cutting-room-management-reduce-fabric-waste.html) and the handover in [cutting room to sewing floor bundle tracking](/blog/cutting-room-to-sewing-floor-bundle-tracking.html).

### Issued to sewn

Between the first operation and the last, pieces are lost to damage during sewing, to bundles that go astray between operations, and to component mismatches where a front panel has no matching back because one bundle went one way and its partner went another.

This is the stage that bundle scanning covers well. The bundle QR carries lot, article, colour, size, component and quantity, and each scan records the operator, machine, operation and timestamp, so a bundle that entered operation four and never reached operation five is a visible fact rather than a rumour. In a factory without scanning this is the hardest stage to audit, because a bundle in a trolley has no record of having existed.

### Sewn to passed

The checking table removes pieces from the shippable count in two ways that people persistently conflate. Some go to rework and come back. Some are downgraded to B-grade and do not.

The second is the one that surprises merchandising. A piece graded B is not scrap — it is a real garment sitting on a real shelf — but it is not available for this shipment. If 120 pieces of the lot graded out, the shippable quantity dropped by 120 while the production report showed 5,000 sewn. Both statements are true and they are describing different things.

### Passed to packed

Even A-grade pieces do not all make it into a carton. Some sit in the repair queue awaiting a fix. Some are pulled as samples by merchandising, QC or a visiting inspector. Some get mixed into a neighbouring lot of the same article in a different colour. And where ratio packing is specified, some are structurally unpackable: if the ratio needs 1,600 mediums and only 1,540 passed, the extra-larges that would have partnered them cannot be packed either, so a shortfall in one size sterilises pieces in the others.

### Packed to dispatched

The smallest stage and still not zero. A carton stays behind in the staging area. A carton is loaded onto the wrong vehicle. The challan is raised for the planned quantity rather than the loaded quantity, so the paperwork and the container disagree from the moment the door closes.

## Why the Two-Number Comparison Fails

Most factories know two numbers: cut quantity and shipped quantity. Subtracting one from the other gives you the size of the problem and nothing else. Five stages can lose pieces and you have one number, so every investigation starts by asking each department whether it was them, and each department correctly answers that its own records balance.

The structural fix is to introduce a third number in the middle, at the boundary between production and finishing. Once you have three numbers you have two gaps, and each gap belongs to an identifiable set of people.

## The Lot Pipeline Ledger

What we run is a pipeline ledger per lot and article with three running balances: **CUT**, **RECEIVED** and **DISPATCHED**.

**CUT** is what the cutting room issued, ideally verified rather than copied off the sheet. **RECEIVED** is what was formally received into finished goods from the sewing floor, recorded by scan with the A-grade and B-grade quantities separated. **DISPATCHED** is what actually left on a challan, built from what was picked rather than from what was planned.

The two gaps then mean specific things. **CUT minus RECEIVED** is the production loss: everything that went wrong between the cutting table and the finishing room door. **RECEIVED minus DISPATCHED** is the finishing and dispatch loss: B-grade sitting on a shelf, pieces in the repair queue, samples pulled, stock stranded by a ratio, cartons not loaded.

That split is most of the value. A production loss and a finishing loss are investigated in different places by different people with different evidence, and before the middle number exists you cannot tell which one you are looking at.

The reason the middle number has to come from a scan rather than a form is that a form is a second job and second jobs get skipped under pressure. Our receipts are created by scanning at the point of handover, numbered atomically as SAC-BSYEAR-NNNN inside a transaction so the sequence cannot duplicate or skip, and each scan shows a large green or red confirmation on screen and announces over the factory PA. The confirmation exists because the quiet failure in any scanning system is the scan that did not register while everyone assumed it did. The broader principle is in the article on [end-to-end tracking without data entry](/blog/end-to-end-garment-tracking-without-data-entry.html).

## How B-Grade and Rework Distort the Count

These two are the most common reason a lot that looks fine on the production report ships short, and they distort in opposite directions.

**B-grade removes pieces from the shippable count without removing them from the produced count.** The garments exist. They were sewn, they were paid for, they appear in the day's output. They simply cannot go in this container. A factory that tracks B-grade only as a total misses the fact that the shortfall is size-specific: if the graded-out pieces are concentrated in one size, the ratio packing problem above is triggered as well. We track B-grade by defect type — stain, tear or hole, size issue, stitching defect — and by lot and article, which is what makes it possible to say how much of the shortfall is recoverable. Stains sometimes clean; holes never do; a size issue is a well-made garment in the wrong size.

**Rework removes pieces temporarily, which is worse, because temporary losses are not budgeted for.** A piece in the repair queue is counted as produced, is not available to pack, and will probably come back — but the question is whether it comes back before the truck leaves. This is why we track the age of the repair queue rather than its size, and why repair bundles get their own QR codes and pay 1.5 times the operation rate: an old repair queue is a leading indicator of a short shipment. The mechanics are in the article on [spot cleaning and rework](/blog/garment-spot-cleaning-rework-flow.html).

Put together, the honest statement of what is available to ship is: received A-grade, minus what is still in the repair queue, minus samples pulled, minus anything stranded by ratio arithmetic. That number is usually smaller than anyone expects on the day before loading, and the whole point of the ledger is to make it visible a fortnight earlier.

## Reconciling at Lot Close

Lot close is the moment to force the numbers to agree, and the discipline is not complicated. For each lot and article, lay out CUT, RECEIVED by grade, DISPATCHED, and then account for the residual explicitly. Every piece in the gap should land in one of a small set of buckets.

- **B-grade in stock**, by defect type — a real, countable, physical quantity sitting on a shelf.
- **In repair** — pieces that exist and are waiting. If this is non-zero at lot close, the lot is not actually closed.
- **Samples pulled** — merchandising, QC, buyer visits, photo shoots. Keep the book.
- **Scrapped** — written off, and recorded as such, because an unrecorded write-off is indistinguishable from a theft.
- **Cutting variance** — the difference between what the cutting sheet claimed and what the room actually issued.
- **Unexplained** — and this bucket should be small and should be reported rather than quietly absorbed. A persistently large unexplained figure is telling you that one of the earlier stages is not being measured.

## What to Check When a Buyer Reports a Shortfall

A checklist, in the order that resolves fastest, because the first three are the ones that most often end the investigation.

1. **Compare the challan to the loading.** Was the challan raised from what was actually picked and loaded, or from the planned quantity? A challan built from the order will always match the order and will not match the container.
2. **Check the carton numbering for gaps.** A gap in the physical sequence is either a lost carton or a numbering error, and they are resolved very differently. Gapless numbering is cheap to enforce and saves this step entirely.
3. **Check for a second challan.** Split shipments are frequently the answer. A partial dispatch went earlier or later and the buyer is comparing against one of them.
4. **Pull the lot ledger.** CUT, RECEIVED, DISPATCHED. If the gap is on the CUT-to-RECEIVED side, it is a production question. If it is RECEIVED-to-DISPATCHED, it is a finishing and packing question. This single step eliminates half the possible causes.
5. **Check the B-grade shelf for that lot and article.** Physically. The quantity is usually right there.
6. **Check the repair queue for that lot.** Pieces waiting on a repair that missed the loading date.
7. **Check for a size-ratio stranding.** If one size fell short under a ratio pack, count how many pieces in the other sizes were sterilised by it.
8. **Check neighbouring lots of the same article.** Mixed lots cancel out in the factory total and show as a shortfall on one purchase order and an overage on another.
9. **Then, and only then, check the cutting record.** It is last on this list not because it is unlikely but because it is the slowest to verify. Do the fast checks first.

## The Honest Limit of the Tracking

That identity ends at folding. Once a garment is pressed, folded and bagged it has no scannable identity of its own, so from packing onwards the honest unit of tracking is the lot and the article. Our receipts, challans and pipeline balances all work at that level.

What this means practically is that a reconciliation can tell you a lot and article is 153 pieces short and which stage lost them. It cannot tell you which specific garment is missing from carton 212, and no CMT factory I know of is applying the per-piece label that would be required to answer that. Anyone selling you piece-level traceability into the carton should be asked exactly which label they are printing and who is applying it.

## Cut-to-Ship as a Standing Number

The last suggestion is to make the ratio of dispatched to cut a number you look at every lot rather than only when a buyer complains. It is a single figure, it is computed from data you already have once the ledger exists, and its movement is more informative than its level. The rest of the operational reporting that sits alongside this is covered in the guide to [WIP tracking](/blog/wip-tracking-garment-factory.html) and the [bundle system](/blog/bundle-system-garment-production-tracking.html) that feeds it.

## Frequently Asked Questions

### Why is the shipped quantity always less than the cut quantity?

Because five separate stages lose pieces and each one loses them differently: cutting issues fewer bundles than the sheet claims, sewing loses pieces to damage and stray bundles, checking downgrades some to B-grade and sends others to rework, packing loses pieces to samples, mixed lots and stock stranded by ratio arithmetic, and dispatch occasionally leaves a carton behind or raises the challan from the planned rather than the loaded quantity. Comparing only the two end numbers tells you the size of the gap and nothing about which stage caused it.

### What is a lot pipeline ledger?

Three running balances per lot and article: CUT, what the cutting room actually issued; RECEIVED, what was formally received into finished goods from sewing with the A and B grade quantities separated; and DISPATCHED, what left on a challan built from what was picked. The two gaps then have distinct owners — CUT minus RECEIVED is a production loss, RECEIVED minus DISPATCHED is a finishing and packing loss. That separation is most of the value, because the two are investigated in different places by different people with different evidence.

### How do B-grade and rework affect the shipped quantity?

They distort it in opposite directions. B-grade permanently removes pieces from the shippable count without removing them from the produced count — the garments exist and were paid for, they simply cannot go in this container. Rework removes pieces temporarily, which is more dangerous because a temporary loss is not planned for: a piece in the repair queue is counted as produced, is not available to pack, and may or may not return before the truck leaves. That is why the age of the repair queue is a better leading indicator of a short shipment than its size.

### What should I check first when a buyer reports a short shipment?

Work fastest first. Compare the challan to what was actually picked and loaded rather than to the order, check the carton numbering for gaps, and check whether a second partial challan exists, because split shipments are a frequent answer. Then pull the lot ledger to see whether the gap sits between CUT and RECEIVED or between RECEIVED and DISPATCHED, which eliminates half the possible causes in one step. After that, physically check the B-grade shelf and the repair queue for that lot, look for stock stranded by a size ratio, and check neighbouring lots of the same article for a mix-up. Verify the cutting record last, because it is the slowest check even though it is a plausible cause.

### Can you track an individual garment all the way to the carton?

Not without a per-piece label that most CMT factories do not apply. Through cutting and sewing the bundle QR carries lot, article, colour, size, component and quantity and every scan records the operator, machine, operation and timestamp, which is genuine piece-level attribution. That identity ends at folding: once a garment is pressed, folded and bagged it has no scannable identity of its own, so from packing onward the honest unit of tracking is the lot and the article. A reconciliation can therefore tell you that a lot and article is short and which stage lost the pieces, but not which specific garment is missing from carton 212.

The reason short shipments feel unsolvable is that most factories are trying to solve them with two numbers taken six weeks apart. Add one honest number in the middle, recorded by scan at the moment goods change hands, and the mystery becomes two much smaller questions that the people involved can actually answer.

*Santosh Rijal is the founder of [Scan ERP](https://scanerp.pro/), a garment manufacturing ERP system designed for factory floor operations. He works directly with sewing lines, cutting rooms, and production supervisors across Nepal's garment manufacturing sector.*