November 6, 2026
A customer sends an artwork update for their reorder: a small QR code tucked in right next to the logo Marketing requested for a campaign they’re running. No big deal. You slug it in, and you run the job the way you’ve run it a hundred times.
Six weeks later, it fails at a self-checkout line a couple of states away, and the reject traces right back to your press.
That little artwork update is the first real change to retail barcode technology since the UPC first graced our boxes. So, as the year winds down, this edition of What the Tech? skips the flashy new operational gadgets to focus instead on preparedness for GS1’s Sunrise 2027 initiative.
Sunrise Who?
Sunrise 2027 is the barcode modernization initiative from GS1, the organization behind global barcode standards. With a deadline that felt comfortably far off, and plenty of other fires to put out, most of us filed it under “later” when it hit our inbox. Sunrise’s goal is to move retail from the one-dimensional UPC to two-dimensional codes (e.g., QR, DataMatrix), read at the point of sale and carrying a much broader set of product data.
By the time the artwork reaches prepress, it looks like just another update from a customer, and the danger is that you treat it like one. The GS1 requirement starts with the retailer, who leans on the brand, who sends you a print file, whose functionality (or lack thereof) at the register could become your latest fire.
By now you’re asking: David, does this even reach a brown-box plant? Yes, and the distinction is case code versus register code. Case codes ship into distribution centers; register codes are the ones a shopper actually scans, and that’s what Sunrise is about. Anything a register has to read is on the table: appliance and fixture cartons, food service and C-store packaging, postprint flexo, etc. If you’re shipping into DCs, you’re already getting deducted for failures (we’ll come back to that). If you’re flexo printing anything on the above list, Sunrise is your problem.
The Code Gets Graded
In the old days, a UPC either scanned or it didn’t, and if we’re honest, the liner was always pretty forgiving. The new standard (3rd Edition, 2024) is ISO/IEC 15415, and it aims to grade several axes at once on a scale from 0 to 4.0: contrast, damage, how square the grid sits, and how far the modules grow or shrink from nominal. GS1’s floor for retail packaging sits at a 1.5 (that’s a low C if you’re trying to relate it to the mental model we all grew up on), and many distributors will be asking you for better. An F won’t scan at all, and a D is really just a little anilox wear away from an F.
Contrast, damage, and out-of-square are all failures we already know how to manage. However, where flexo print will really start to show its cracks is along the growth axis (that’s dot gain for us lab-coat-wearing types). A two-dimensional code can experience 15%–20% print growth that will cause squares to bleed together and the grid to become unscannable. Error correction will cover some of that loss, but when it runs out, the code goes dark. The answer to much of this is good old-fashioned prepress work. Unfortunately, that work rarely gets done when a code shows up and leaves as an “art update from marketing.”
Verify It
Most independents I talk to follow the ISO 9001 standard, if they carry any certification at all. Unfortunately, it tells you that your process was right but doesn’t tell you whether a code actually works. “That’s what a scanner is for, right?” Well, sort of.
A scanner answers one question: Does it read, yes or no? You pull the trigger, it beeps, and the job moves on. A verifier does something else entirely. It grades the symbol against the 15415 measures, tells you how close to the line you’re walking, and is one tool we should all plan to add. A decent one runs $3,500 to $5,000, and one gun won’t cover a multishift operation, so budget for tooling up.
Retailers already inspect supplier barcodes and deduct for the failures. Walmart’s Supplier Quality Excellence Program does it on their case codes (the GTIN-14 its DCs scan at receiving) today, for example. The per-incident hit is small, a couple hundred in admin fees plus a capped per-case penalty, so you’d typically just pay it and move on. The bigger issue, as many of you already know, is the scorecard that’s used to track repeated defects and decide who gets the allocation during category review. The register code is next in line for this. When it fails, your customer’s product won’t scan, they’ll trace it to your press, and the plant across town will be glad to say they would have done it better. Those verifier guns act as insurance against the fine, but more importantly, against being the plant that loses a customer over a QR code.
It Starts With Static
Getting it right today is straightforward, but that also means there’s really no alibi for getting it wrong. The codes showing up right now are static, the same on every box of a SKU, they carry the GTIN and drop into the plates you already run. Static is the part we can knock out in Q1. GS1 hands out a size estimator and a retail implementation guide for free. Start there.
What staying ahead requires is heightened prepress discipline and getting a verifier added to your QC toolkit. With just those in place today, the grade is something you can own before you ship, and you’ll be more prepared as things evolve. But keep an eye out for the pretty ones. When Marketing sends a stylized QR with rounded corners and brand colors, that styling is the thing that will drag your grade down.
With the amount of data that can be packed into a 2D code, someday a customer is going to want a unique code per box. That will turn the discussion toward inline inkjet, case coders, or label machines, but that’s a different conversation for a different column.
So, the next time new artwork lands with a quiet QR sitting on it, treat it like the press spec it is, and grade it before it ships. It’s getting graded either way. The other grader is a register two states over, reading your print whether you’ve checked it or not.
David Wiens is CEO of BPS AI Software. He can be reached at david@bpsaisoftware.com.
