GS1 Sunrise 2027: what the 2D barcode switch means for your store

the most successful piece of technology in your store is fifty years old

The UPC barcode was scanned for the first time at a supermarket checkout in 1974, and the most remarkable thing about it is not that it worked — it's that nothing has replaced it since. Fifty years of near-perfect reliability, effectively zero marginal cost, printed on everything from a pack of gum to a pallet of motor oil. There is no other piece of retail technology with a record like that. Your POS system has been replaced four times. Your scales get recalibrated annually. The barcode just keeps working.

It keeps working because it does exactly one thing. A UPC carries a single fact: what this product is. Not when it expires. Not which batch it came from. Not which individual unit it is. One number, the GTIN, which your POS looks up to find a price. Everything else a modern store needs to know about a product — the expiry date, the lot number, where it came from, whether it's part of a recall — lives somewhere outside the barcode. On the pack in human-readable ink, in a supplier's despatch paperwork, in somebody's memory, or in a field a staff member typed in by hand at the receiving bench.

Sunrise 2027 is the retail industry's coordinated plan to fix that. It's worth understanding precisely, because the gap between what it actually is and what it sounds like it is — a deadline, a mandate, a thing you can be fined for missing — is wide, and a lot of the commentary aimed at store owners lives in that gap.

what Sunrise 2027 actually is (and is not)

Sunrise 2027 is an initiative led by GS1, the standards organization that administers the barcode system. The ambition is specific: by the end of 2027, retail point-of-sale systems should be capable of scanning 2D barcodes — meaning a QR code carrying a GS1 Digital Link URI, or a GS1 DataMatrix — in addition to the 1D barcodes they scan today.

Read that sentence again, because every word is doing work. It's a capability milestone, not a mandate. There is no law behind it, no regulator, no fine for non-compliance. Nobody is going to inspect your checkout in January 2028. It is also not a ban on 1D barcodes — the UPC is not being switched off, and during the transition, products that carry a 2D code also carry the familiar 1D barcode right alongside it, so a store that scans nothing but UPCs in 2028 will still ring up sales just fine.

GS1 has run this play before. The 2005 Sunrise asked US retailers to be able to scan 13-digit EAN barcodes alongside 12-digit UPCs, and it worked the same way: a coordinated date that gave manufacturers, hardware vendors, and retailers a common target so the chicken-and-egg problem — nobody prints the new code because nobody can scan it, nobody can scan it because nobody prints it — resolves itself. That's the real function of Sunrise 2027. It's not a deadline aimed at you. It's a synchronization signal aimed at an entire industry, and you're one of the parties being synchronized.

So why should a store owner care about a voluntary industry milestone? Because the milestone marks the moment the interesting data starts arriving on packs in machine-readable form, and the stores equipped to read it get capabilities the stores that aren't simply don't have.

what a 2D barcode actually carries

A 2D barcode — the square, pixelated kind, rather than the vertical bars — carries dramatically more data in less label space. What matters for retail isn't the raw capacity, though. It's the structure. GS1 2D barcodes encode data as a chain of Application Identifiers, or AIs: short numeric prefixes that announce what the next chunk of data means. The ones a store owner will meet daily:

AI (01) — GTIN. The product identity. Same number the UPC carries, so the price lookup works the same way.

AI (10) — batch or lot number. Which production run this pack came from. This is the field every recall is keyed to.

AI (17) — expiration date. The use-by date, in a fixed machine-readable format. No ambiguity about whether 03/07 means March or July.

AI (21) — serial number. A number unique to this individual pack, not just this product. Pharmacy already lives on this one.

AI (15) — best-before. The quality date, distinct from the safety date, for categories where that distinction matters.

So where a UPC says "this is a 500g tub of yogurt," a GS1 DataMatrix or GS1 Digital Link QR on the same tub can say "this is a 500g tub of yogurt, from batch L2417, expiring June 28, serial number such-and-such." One scan, four facts, zero typing. If you've ever stood at a receiving bench keying batch numbers and expiry dates off invoices line by line, you already understand why this is the part of Sunrise 2027 worth getting excited about — and we've written a whole piece on exactly that workflow change.

The two symbologies are functionally siblings. GS1 DataMatrix is the compact industrial one — it's what pharma uses. The GS1 Digital Link QR encodes the same identifiers inside a web address, so the same square serves your scanner and a shopper's phone camera, which resolves to a product page. The differences, and which scanners read what, are covered in our store owner's guide to the three symbologies.

what changes at the point of sale

Less than you'd think, and the part that does change is software, not ceremony. The checkout transaction is the same: scan, look up GTIN, price, total. A 2D-capable scanner reading a GS1 barcode still produces a GTIN — it just arrives wrapped in that AI structure along with the other fields.

And that wrapping is where unprepared POS systems faceplant. The classic failure mode: the scanner dutifully reads the DataMatrix, transmits the full AI string, and the POS — expecting a bare 12-digit UPC — dumps the entire string into its product lookup, finds no match, and beeps the unhappy beep. The hardware worked perfectly. The software didn't know how to take the GTIN out of the envelope. "2D-capable POS" means software that parses AIs, not just a scanner with a camera in it.

Once the POS does parse AIs, genuinely new things become possible at the till. The expiry date is in the scan, which means the POS can refuse to sell an expired item — not as a policy in a training binder, but as a hard stop at the register. Markdowns can key off the actual date on the actual pack rather than a shelf-edge guess. Recalled batch numbers can be blocked at checkout the hour the recall notice lands. None of this is science fiction; all of it is arithmetic on data that's now in the beep.

what changes at receiving — which is the bigger story

The POS gets the headlines because Sunrise 2027 is defined as a POS capability milestone. But the back door is where the money is.

Receiving is where batch and expiry data enters your system today, and in most independent stores it enters through someone's fingertips — read the date off the pack or the invoice, type it in, next line, forty more to go. Every downstream process that depends on that data (FEFO rotation, expiry alerts, markdown timing, recall response) is only as good as that typing. A 2D barcode at receiving replaces the whole exercise with a scan per line: product, batch, and expiry captured together, exactly as printed, in under a second.

Pharmacy is the proof this works at national scale, because US pharmacy didn't wait for a sunrise — it got a statute. Under DSCSA, every salable pharmaceutical unit already carries a GS1 DataMatrix encoding the NDC/GTIN, serial number, lot, and expiry, and serialization enforcement is now live across the chain: manufacturers since May 2025, wholesalers since August 2025, large dispensers since November 2025, with small dispensers exempt only until November 27, 2026. Every independent pharmacy receiving stock today is living in the post-1D world the rest of retail is walking toward — and the DSCSA compliance checklist shows what the receiving discipline looks like once package-level data capture is non-negotiable.

Grocery has its own regulatory tailwind. FSMA 204, the FDA's food traceability rule, requires lot-level records for foods on the traceability list, with enforcement now set for July 20, 2028 — conveniently just after Sunrise 2027 puts lot numbers on packs in scannable form. That timing is not a coincidence the industry is embarrassed about. If you sell foods on the traceability list, the FSMA 204 compliance picture is worth reading next to this one, because the 2D barcode is the cheapest way the required data will ever arrive at your back door.

And for the years when many packs still carry only a 1D barcode — which is most of the transition — the same batch and expiry data can arrive digitally in supplier despatch files instead, which is why a supplier file receiving workflow belongs in the same conversation.

what to do this year

Not in 2027. This year, 2026, while there's no pressure and the moves are cheap.

Audit your scanner hardware. Pick up each scanner in the building and check whether it's a laser or a camera-based imager. Lasers physically cannot read 2D codes — they sweep a single line of light and a DataMatrix is invisible to them. Imagers read everything. The good news: most USB scanner guns sold today are 2D-capable imagers, so any scanner bought in the last few years is probably fine, and any scanner you replace from here forward should simply never be a 1D-only unit. The model-number spec sheet settles it in thirty seconds.

Ask your POS vendor one precise question. Not "are you ready for Sunrise 2027" — they'll say yes. Ask: "when a scanner sends you a full GS1 AI string from a DataMatrix, do you parse out the GTIN and the other identifiers, or does the lookup fail?" The quality of the answer tells you most of what you need to know.

Ask your key suppliers about their 2D printing plans. Especially private-label and fresh suppliers, where you have leverage and where expiry data matters most. You're not demanding anything; you're finding out which of your categories will go 2D first, so you can pilot receiving-by-scan where it'll pay off earliest.

Choose inventory software that parses AIs. This is the load-bearing decision. The scanner is a commodity; the intelligence lives in what the software does with the scan — splitting the AI string into product, batch, expiry, and serial, and filing each where it belongs. Software that treats a 2D scan as a long weird product code gives you nothing the UPC didn't.

Start at receiving, not at the till. POS capability is the industry's milestone; receiving capability is your payback. You don't need a single customer-facing change to start capturing batch and expiry by scan on the packs that already carry 2D codes today.

what not to panic about

There is no deadline with teeth. Nothing happens to a store that scans only 1D barcodes on January 1, 2028. The initiative is a coordination point, not a compliance event.

The UPC isn't disappearing. Dual-marking — 1D and 2D side by side — is the explicit transition plan. Your existing checkout keeps working throughout.

You don't need to replace your POS. You need to ask it one question (see above) and plan upgrades with 2D in mind, the way you'd plan any normal refresh cycle.

Every pack will not be 2D-labeled in 2027. Adoption will be uneven for years — pharma is already done, fresh food will move with FSMA 204, long-tail dry goods will lag. Build a workflow that handles both, because both is what your receiving bench will see for a long time.

The honest summary: Sunrise 2027 asks almost nothing of a small retailer and offers quite a lot. The barcode that's been telling you what for fifty years is learning to tell you which one and until when. The stores that wire those two new facts into receiving, rotation, and markdowns will quietly waste less and recall faster than the ones still typing — and the price of entry is a question to your POS vendor and a scanner you probably already own.

ShelfLifePro reads GS1 DataMatrix and GS1 Digital Link QR codes out of the box — one scan at receiving captures product, batch, and expiry, and feeds FEFO rotation and expiry alerts automatically. Ready for Sunrise 2027 without waiting for it.