The Shipping Label Guide for European E-Commerce: Formats, Barcodes and the Errors That Cost You Deliveries

A shipping label looks like the least interesting part of the fulfillment process. It is the thing a warehouse worker sticks on a box thirty seconds before it disappears onto a truck. But that small rectangle of paper carries the entire routing intelligence a parcel network relies on: the destination, the carrier's sort code, the service level, the customs data, and above all, the barcode that every scanner between your warehouse and the customer's doorstep will read. When that barcode fails to scan, the parcel does not just sit there quietly. It gets kicked out of the automated sortation line, routed to a manual re-key station, delayed by hours or days, and in the worst cases, lost entirely.

At Zineps we sit on top of hundreds of thousands of label generations a month across DHL, DPD, PostNL, GLS, UPS, bpost and dozens of regional carriers. We see, in aggregate, which label mistakes actually cause delivery failures and which ones are harmless. This guide is written from that vantage point: not as a definition of what a shipping label is, but as an operational breakdown of how labels work, why they break, and how to build a label process for a growing e-commerce business that never has to think about barcodes again.

What Is Actually Encoded on a Shipping Label

A shipping label carries two parallel layers of information: a machine-readable layer and a human-readable layer. Carriers optimize almost entirely for the first one.

The Barcode Layer

Most parcel carriers still use linear barcodes. Code 128 is the most common symbology for parcel tracking numbers because it is dense enough to hold an eighteen to twenty digit tracking number in a compact space and read reliably at conveyor belt speed. Freight and pallet labels, and increasingly multi-parcel shipments in Europe, use GS1-128, a structured variant of Code 128 governed by GS1, a global standards body, not a carrier, that wraps data in Application Identifiers. The most important one is AI (00), the Serial Shipping Container Code, or SSCC, an eighteen-digit number that uniquely identifies a single logistics unit anywhere in the world. If you have ever wondered why a pallet or a consolidated multi-box shipment needs its own label separate from each parcel's, the SSCC is the reason: it lets the carrier scan one code and know it represents a specific, traceable unit rather than an assumption based on address matching.

QR codes are gradually being added alongside linear barcodes, mostly to support carrier self-service returns and out-of-home tracking at pickup points, but they remain a secondary code across Europe, not a replacement for Code 128 or GS1-128.

The Human-Readable Layer

Underneath the barcode sits the information a person can act on if the scanner fails: the destination address, the sender's return address, the service type, the tracking number in plain digits, and often a routing code specific to that carrier's regional sort facility. This is the layer manual re-key stations rely on when a barcode does not scan, and it is also the layer that gets a parcel misrouted when it is wrong even though the barcode itself is perfectly fine. This is a mistake we see constantly at Zineps: a merchant's own address book has a typo in it, the barcode encodes the tracking number correctly, but the human-readable address printed underneath sends the parcel to the wrong sort facility the moment a scanner has to fall back to manual handling.

Why the 4x6 Format and GS1-128 Standard Exist

Every major European and global carrier, DHL, UPS, FedEx, DPD, GLS, PostNL, has converged on a four by six inch label, roughly ten by fifteen centimeters, as the practical standard, printed at a minimum of 203 DPI. This is not an arbitrary choice. At that size, a barcode can be printed large enough and with enough contrast to remain scannable even when the label picks up a light crease, a smear of packing tape, or a bit of warehouse dust, all of which happen constantly at real-world volumes. According to the official GS1 Logistic Label Guideline, print quality for a GS1-128 barcode must meet a minimum grade of 1.5 under ISO/IEC 15416, a standard most office inkjet printers cannot reliably hit, which is one of the quieter reasons thermal printing dominates parcel shipping.

Thermal printing itself comes in two forms. Direct thermal uses heat-sensitive label stock that darkens where the print head touches it, and is what most carriers ship on by default because it is cheap and fast. Thermal transfer uses a ribbon to melt ink onto a more durable label material and is preferred for anything that needs to survive months in transit or extreme temperatures, such as cold chain or industrial freight. For a typical e-commerce parcel moving from a Dutch warehouse to a German or French customer, direct thermal is fine. For anything crossing multiple customs zones with a longer transit time, thermal transfer labels resist fading far better.

The Real Cost of a Bad Label

It is tempting to treat a mislabeled parcel as a minor inconvenience. The data says otherwise. Independent estimates put the barcode legibility failure rate at roughly one in every five hundred parcels across major carrier networks, which sounds small until the math is run against real volume. A merchant shipping twenty thousand parcels a month is statistically generating around forty unreadable labels a month, each of which triggers one of three outcomes: a manual re-key at a sort facility that adds a full day or more to transit time, a routing exception that requires customer service intervention, or, in a small but real percentage of cases, a lost parcel that becomes a refund, a replacement shipment, and a damaged customer relationship all at once.

A rejected label sits at the intersection of two problems we write about often: delivery failures that erode customer trust, and hidden carrier costs that quietly eat into margin. Most carriers charge a manual handling fee on top of the delay when their sort facility has to re-key a shipment by hand, which means a bad label is rarely just a delay. It is also a line item.

The Five Label Errors We See Most Often in E-Commerce

• Wrong label size or DPI for the target carrier. A label generated at a lower resolution for one carrier's API can print as an unreadable blur on a different carrier's thermal printer if formats are not translated automatically per carrier.
• ZPL sent where a PDF was expected, or the reverse. Thermal printers speak ZPL, Zebra Programming Language, natively, while many office and hybrid printers only accept PDF or PNG. Running more than one fulfillment location almost always means running more than one printer type.
• Missing SSCC on multi-parcel or palletized shipments. When an order ships as more than one box, carriers expect each box labeled individually and, for larger consolidated shipments, an SSCC tying them together. Skip it and the carrier's system cannot confirm the shipment is complete, a common and invisible cause of partial delivery disputes.
• Tape over the barcode. This sounds almost too simple to mention, but it remains one of the single largest causes of manual re-keys at high-speed sort facilities, because clear packing tape reflects light in a way that confuses laser and camera-based scanners at conveyor speed.
• Stale or duplicate labels from reused packaging or reprinted batches. When a return box gets reused, or a warehouse operator reprints a batch after a system hiccup without voiding the first set, two valid barcodes end up on the same parcel, and the carrier scans whichever one is more legible, which can route the shipment under the wrong tracking number entirely.

Why This Gets Harder, Not Easier, as You Add Carriers

A single-carrier operation can survive with a manually maintained label template, because there is only one specification to get right. The moment a business adds a second and third carrier to improve delivery coverage and reduce cost per shipment, a practice we explored in detail in our guide to multi-carrier shipping strategy, it also inherits a second and third label specification, each with its own size tolerance, barcode symbology preference, required fields, and API quirks. The label layer is the part of that multi-carrier resilience case that gets the least attention and causes the most day-to-day friction, precisely because it stays invisible until something breaks.

This is exactly the layer where manually maintained shipping software starts to crack. Spreadsheet-driven address books, hardcoded printer templates, and carrier integrations built one at a time by different developers in different years tend to produce exactly the kind of inconsistency described above: correct most of the time, and wrong in ways nobody notices until a customer complains.

How Zineps Treats Labels as Infrastructure, Not a Print Button

This is precisely the problem Zineps was built to remove. As the operating system for shipments, Zineps sits between your order management system and every carrier you use, and normalizes label generation into a single, consistent layer regardless of which carrier ends up fulfilling a given order.

In practice that means three things. First, address and format validation happens before a label is generated, not after a carrier rejects it, catching the kind of address typo or missing field that would otherwise slip through to the human-readable layer of a label undetected. Second, every label is generated in the exact format, size, DPI, and symbology each specific carrier requires, whether that is a DHL DE label, a GLS file in ZPL, or a PostNL label in PDF, without a warehouse team needing to know or manage the difference. Third, because Zineps sees label generation and delivery outcomes across the same shipment, it can surface which carriers, which printer stations, or which SKUs, oversized or irregularly shaped items are disproportionately prone to label damage, are generating a disproportionate share of scan failures, turning an invisible cost into a measurable, fixable operational metric.

For a growing e-commerce or fulfillment business, this is the difference between treating a shipping label as a static template someone configured two years ago, and treating it as a live, monitored part of your logistics infrastructure, the same way you would monitor checkout conversion or cart abandonment. You can see how this fits into the wider Zineps shipping platform here.

A Quick Checklist Before Your Next Print Run

• Confirm your label size and DPI settings match each carrier's current minimum, not just the one you originally integrated with.
• Separate ZPL and PDF output by printer station so warehouse staff are never guessing which template applies.
• Add SSCC generation automatically for any shipment leaving as more than one parcel.
• Ban clear tape over the barcode zone as a packing station rule, not a suggestion.
• Track scan failure and manual re-key rate as a monthly metric, the same way you already track shipping cost per order.
• Automate address validation ahead of label creation rather than relying on manual review.

The Bottom Line

A shipping label is the smallest, cheapest part of the fulfillment process, and it is also the single point where a mistake most reliably turns into a delayed, disputed, or lost delivery. European e-commerce businesses spend enormous energy optimizing checkout, negotiating carrier rates, and building loyalty programs, while the physical artifact that actually gets the product to the customer often still runs on whatever template was set up when the company shipped its first hundred orders. Fixing that does not require a new carrier contract or a pricing renegotiation. It requires treating label generation as infrastructure, validated, standardized, and monitored, rather than an afterthought bolted onto an order confirmation email.

That is the layer Zineps was built to own. Explore Zineps pricing and see how a unified shipping layer fits your carrier mix.