Written by Marcus Rivera — Former Amazon Warehouse Manager. More about me →
I used to think generating these was straightforward — pick a type, enter the data, export. Then I had a situation that changed my mind. I produced a batch, shipped it, and the recipient's system rejected every single item. Nothing looked wrong on my screen.
The root cause was not the tool. It was that I did not understand what the destination actually needed. GS1 General Specifications spells out the requirements clearly, but most generators do not enforce them. That gap — between what looks right and what actually works — is where most problems live.
The experience reshaped how I work. Now I treat every generation as a two-step process: configure for the destination, then verify against the real reader. It added five minutes per job and removed an entire category of failures.
Your screen renders everything at 72-96 DPI with full color and anti-aliasing. The real world does not. A scanner reads at 200-300 DPI with a laser that needs clean edges and clear contrast. An icon might render at 16 pixels on one device and 180 pixels on another — same file, completely different rendering engines.
I have debugged production failures that did not reproduce on any of my devices. The issue was always the same: the output met the spec on paper but failed under real-world conditions that GS1 General Specifications did not anticipate. The spec defines the minimum; real-world usage demands more.
That gap between spec and reality is permanent. The only defense is to test at the actual size, with the actual reader, and in the actual environment — not a simulated version of it.
People search for a generator, pick the first result, type something in, and hit download. That is fine if you are making a one-off label for a personal project. It is a disaster if you are printing labels for a supply chain or deploying icons to a production website.
Every type exists for a reason. GS1 General Specifications defines distinct use cases, and using the wrong type for your scenario is worse than using no type at all — because a wrong type that scans successfully gives you false confidence. I have seen shipments get rejected at receiving docks because someone used a generic format when the receiver required a registered one.
The cost of guessing wrong is rarely visible on the day you generate the file. It shows up weeks later, when a scanner refuses to read a label or an iPhone shows a blank square. That lag is why so many people make the same mistake once.
At the end of the day, the goal is an output you do not have to worry about. If a single step here saves you one redo, it was worth the read. I keep ISO/IEC 15417:2007 for Code 128 symbology bookmarked for the days I doubt myself, and I run my checks on every export before it ships.
Written by Marcus Rivera — Former Amazon Warehouse Manager. More about me →