One morning this lands in your support inbox: "My order arrived but it's the wrong item."
You check the tracking: everything looks fine. The parcel left on time, arrived undamaged, and the recipient signed. The carrier made no mistake. The mistake happened days earlier, in the warehouse or at your packing table, in a matter of minutes.
What makes this error so annoying is that no system tells you about it in advance. You see a delay on the tracking screen, and the carrier flags a bad address. A wrong item, though, only the customer finds.
This guide covers the real cost of shipping the wrong item, its three distinct types, and a practical way to catch it before the label is printed. It also covers where that method stops working.
This post focuses on the part up to the moment the parcel leaves. For how to handle the complaint after a faulty item, see our delivery complaints guide; for who should pay the return shipping, see the return shipping cost guide.
Shipping the Wrong Item Isn't a Carrier Error, It's Its Own Class of Error
The industry calls this a misshipment: the wrong item packed, the right item sent to the wrong customer, an extra product added, or the same order shipped twice. What sets it apart from loss and delay is that the carrier did its job correctly; the problem came from inside.
Two consequences follow:
- Carrier tracking data never shows this error. Metrics like delivery time and on-time rate can look spotless while a customer is unhappy.
- You can't pass it on to the carrier. Compensation or switching carriers won't fix it; the fix is in your own process.
The Real Cost of the Mistake
It's easy to think of a wrong shipment as "one more parcel." In reality there are more line items:
| Line item | What happens |
|---|---|
| Three shipments | The wrong item going out, coming back, and the right item going out again. You needed one |
| Product risk | The wrong item often comes back opened or damaged, and sometimes doesn't come back at all |
| Labor | A second pick, a second pack, a second label |
| Support time | Back-and-forth, return instructions, follow-up |
| The customer | A low rating, a bad review, a purchase that doesn't repeat |
A small example: say the average shipping cost is 100 (put in your own number). One mistake means 200 of pure extra shipping. Ship 1,000 parcels a month, get five in a thousand wrong, and that's five mistakes a month and 1,000 in shipping alone. On its own that may not look big. The real loss is whether those five customers ever order again.
There is some measure of this on the retail side. In a survey cited by Access Group (over 8,000 e-commerce customers, sourced to Multichannel Merchant), 73% of respondents said they had experienced a late or incorrect delivery, and roughly a quarter of those said it led them to cut ties with the retailer. The figure covers late deliveries too, not just wrong items, so don't read it as a wrong-item rate. But the direction is clear: a faulty delivery is a risk of quietly losing the customer who got it.
There is one thing you can do today to find your own rate: make "wrong item" its own category in return and complaint reasons. Most merchants don't know this number, because "wrong item" usually ends up inside "other." The shipping KPIs guide likewise recommends tracking reasons so you can separate carrier-caused returns (damage, wrong item) from product-caused ones.
Three Different Errors, Three Different Places to Catch Them
"The wrong item went out" isn't one mistake. Depending on how it happens, it is caught in a different place.
| Error | How it happens | Where it's caught |
|---|---|---|
| Wrong item or variant | Navy instead of blue, L instead of M, two similar boxes | At picking or packing, by comparing against the SKU |
| Missing or extra quantity | Two instead of three, three instead of two | At packing, by counting per line |
| Mix-up between orders | A's item ended up in B's box; totals are right, placement is wrong | Only at the packing bench, order by order |
The third row is the sneakiest. When you batch-pick, collecting items for many orders in one trip into the same crate, the item count in the crate comes out right. But one item may have gone into the wrong box, and no total will show it. That is why two measures are applied together: scan-based pick-and-pack and a final scan before the label. One catches the shelf error, the other the box error.
Industry guides to shipping errors point to the same two places: look-alike items getting swapped, and package contents not being checked against the order before sealing. For fixes they list barcode scanners alongside warehouse investments like pick-to-light and voice picking. At 50-5,000 shipments a month those investments are often unnecessary; the first two, scanning and a final check, already close most of the gap.
Why Your Current Method Isn't Enough
Most merchants do this by eye: look at the order screen or printout and put the item in the box. That works at 10-20 orders a day. As order counts grow, three things change:
- Attention leaks. Someone who has done the same motion hundreds of times no longer reads two similar items, they recognize them. The error happens at exactly that moment of "recognition."
- Look-alike products multiply. As the catalog grows, size, color and model variants converge.
- Batch picking becomes common. Picking by product to save time makes sense, but it brings the third type of error above.
As volume grows, checking shrinks, because a check is an extra step and the first thing skipped when things get busy. As we noted in our guide to scaling shipping operations, in a manual process wrong shipments, address errors and forgotten orders start at around 30 orders.
Scanning Before the Label: How It Works
The idea is simple: the item is compared against the order before the label is printed. The system does the comparing, not a person; the person just scans the item's barcode.
In Shipink this exists in two places.
1. The bulk picking list. When you select several orders on the Orders screen and open the picking list, the orders' items are grouped by product: you see the total quantity needed per product and how many orders it appears in. As you scan items off the shelf, the count goes up. Scan an item that isn't on the list and you get a warning; scan more than the required quantity and you get a separate one. You can print the list grouped by order or by product; on a half-finished pick, the printed list also shows how many units have been picked.
2. Scanning while creating a shipment. When you open "Scan items" while creating an order's shipment, every item you scan is matched against that order's lines. Scan an item that isn't on the order and you get a red warning and an audible beep, so the person at the bench hears it even without looking at the screen. When every line is complete, an "All scanned" notice appears. On sales channels that support partial shipments, what you scan goes into the shipment: whatever you scanned is what ships.
The two steps do different jobs. In the bulk list items are counted per product, not per order, so it can't know which order an item belongs to; it catches shelf errors. Scanning on the shipment screen verifies a single order, so it catches the third error, the box mix-up. If you batch-pick, use both together.
The flow is independent of the sales channel: whichever sales channel an order came from, it is verified on the same screen, so you don't need to run a separate item check in each channel's own panel. We covered why that matters in the multichannel order management guide.
Honest Limits
This method doesn't close every error. What you should know:
- Matching is by SKU. The code you scan has to equal the SKU of the item on the order. If the barcode on the product (a manufacturer's EAN, for example) differs from your SKU, it won't match. In that case you need to print your own SKU label for the product.
- Items without a SKU can't be scanned. When such a line exists, matching doesn't work and you tick the line off by hand. If your catalog has items with no SKU, or variants sharing the same SKU, fix those first; otherwise the system can't tell the variants apart.
- It warns, it doesn't block. Scanning doesn't lock label creation. You can see the wrong item and carry on anyway; enforcing the check is your process decision.
- Scans aren't recorded. There's no audit trail of who verified which order at what time.
- It doesn't track stock. Shipink doesn't count inventory; your sales channel owns your stock. Scanning only answers "is this item on this order?" It doesn't solve out-of-stock items or shelf layout.
- No camera scanning. You use a barcode scanner or type the SKU by hand. The scanner needs no special software; a standard device that behaves like a keyboard is enough.
These limits also show the method's purpose: to catch what a person's eye misses at the last point before the label, without building a warehouse management system. If you need full warehouse management, this doesn't replace it.
Three Clean-Ups Before You Start Scanning
Scanning depends on clean data. Within a week:
- Give every variant its own single SKU. If blue M and blue L share a SKU, neither the system nor you can tell them apart.
- Find the items with no SKU. Go through your product list and fill in the blank SKUs one by one.
- Separate look-alike products on the shelf. Don't put two near-identical items side by side. Scanning catches the mistake; the layout of the shelf helps it not happen at all.
Why Now?
The November campaigns are coming. In peak season two things happen together: order volume goes up, and packing usually takes on extra or temporary people. If you try to set up the process that week, learning and mistakes happen at the same time. In the peak season preparation plan we described how exceptions grow with volume and the cost lands on you; a wrong item is one of those exceptions.
A mistake in the campaign comes back as a return at year-end. A wrong item sent in November reaches you in December as a return or complaint, when customer patience and support capacity are at their lowest.
SKU clean-up takes time. Filling in the missing SKUs in a catalog isn't an afternoon's job. With a few weeks to go it's comfortable; in campaign week it isn't possible.
A Rehearsal You Can Run This Week
- Go through the last 30 days of returns and complaints. How many were wrong items? If there's no category, create one from now on.
- List the items with missing or duplicate SKUs and fix them.
- Get a basic barcode scanner and make sure products carry a label that contains the SKU.
- Run a 20-order rehearsal. First select those orders in bulk and scan from the picking list. Then scan each order one by one while creating its shipment. Scan a wrong item on purpose and check the warning.
- Time it. See how much extra time each order takes. The extra time is usually small next to the cost of one mistake.
Don't Let the Customer Find It First
A wrong shipment is one of the costliest process errors: invisible in tracking, first reported by the customer, and often ending with a customer who won't buy again. Putting the check at the last point before the label shows you the mistake before the customer finds it.
If you want to verify orders by scanning and see the picking list, you can try Shipink for free, look at the operations tools, or write to us to talk through your setup.
