For the VP of Store Systems staring down a lane estate measured in thousands, here is the entire obligation, quoted at its actual size: GS1's Sunrise 2027 programme expects retail point-of-sale to scan and process 2D barcodes — extracting the GTIN — by the end of December 2027. That is GS1's programme, published as GS1's 2D-in-retail transition, and it is theirs: an industry milestone, not a law, not a regulation, not a mandate with an enforcement mechanism. Everything you have been told beyond that sentence is someone's sales pitch, and this brief exists so you can tell which parts.
Decomposing "scan and process"
The one-sentence expectation unpacks into exactly four lane capabilities. Each is a separately testable, separately procurable thing:
| Capability | What it means at the lane | Where it lives |
|---|---|---|
| 1. Read the symbol | The imager decodes 2D symbologies — QR and GS1 DataMatrix — not just 1D linear | Scanner hardware / firmware |
| 2. Parse the syntax | The decoded payload is parsed — GS1 element strings or a GS1 Digital Link URI, both must parse | POS software / scanner firmware |
| 3. Extract the GTIN | The GTIN comes out of either syntax and hits the item lookup exactly as a UPC scan does today | POS software |
| 4. Transact | The sale rings — including the transition years when one pack carries both a 1D and a 2D mark, and the lane must not double-scan or mis-ring | POS software + lane procedure |
That is the whole of Sunrise at a lane. Capability 1 is mostly bought already — modern imagers read 2D. Capabilities 2 through 4 are software versions and rollout discipline. The scanner-replacement question specifically is worked in do we have to replace POS scanners for 2D by 2027, and the capital-scheduling math is its own decision — multi-quarter, backward from December 2027.
The six corrections a GS1-literate reader will catch vendors on
Carry this list into every vendor meeting. Each item is a sentence that, if a vendor says it, tells you they are selling past their knowledge:
- "hasGS1DigitalLink went core in schema.org v30.0." It did not — it carries pending.schema.org part-of status.
- "hasGS1DigitalLink is new." It has shipped since schema.org v27.0.
- "A DataMatrix or QR is scannable by a consumer's phone." GS1 DataMatrix clears the lane but is not reliably decoded by native phone camera apps; QR is the consumer-scannable symbology. A vendor who conflates them has failed the cheapest diligence question in the programme.
- "The 2D code links to a web page." Only if it carries a GS1 Digital Link URI. A 2D symbol carrying GS1 element strings (FNC1 syntax) is not a URI and resolves to nothing in a browser. Your lane must parse both; which one is on the pack is the brand's choice, not yours and not GS1's.
- "Sunrise means lot and expiry come to POS." Sunrise's baseline is GTIN extraction. Lot and expiry are additional AIs a brand may choose to encode. Some will; the lane must tolerate their presence; nothing obliges their existence.
- "The 2D mark Sunrise puts on packs." Sunrise puts nothing on anyone's pack. The brand does, if it chooses — the mark decisions belong to the GTIN owner, which is precisely why your suppliers' packs will arrive in mixed states for years. (The GTIN itself is worked engineer-literal at barcoding.dev, by visibility.cloud.)
Strike from any vendor deck: consumer-engagement promises attached to lane hardware, "compliance" language implying legal force, and any timeline that treats December 2027 as your suppliers' re-marking deadline rather than GS1's POS-capability milestone.
What Sunrise does not require of you
Worth stating as a list, because the padding in vendor proposals is made of these: Sunrise 2027 does not require you to print anything, to resolve Digital Link URIs to the open web at the lane, to capture lot or expiry at checkout, to build consumer experiences, to replace your loyalty stack, or to stand up any traceability system. A proposal that bundles those into "Sunrise readiness" is bundling its margin, not your obligation.
The question the programme leaves open — and it is the expensive one
Here is what the four capabilities above conspicuously do not include: anywhere for what the lane reads to go. Sunrise makes your lanes readers of a richer mark. When a supplier does choose to encode lot and expiry, your refreshed lane will decode data your transaction log has no field for — the POS transaction answers what sold, never what happened to this item. And your organization sits on the other clock too: the same covered foods crossing your lanes cross your receiving docks under FSMA 204, where the record question has real dates attached — scoped in the fresh-department guide.
The event record — where the richer mark's data lands, joined at the same GTIN your lane extracts — is a decision adjacent to the POS refresh, not inside it, and it is cheapest made while the refresh capital is already moving. That full two-clock read for grocery is the pillar: Sunrise 2027 and FSMA 204 for grocery.
Sunrise, correctly sized, is smaller than the pitches and earlier than the calendar suggests: four lane capabilities, scheduled backward from December 2027, plus one adjacent decision the programme leaves entirely to you.
Where this goes next
visibility.cloud provisions capture workspaces from the seat list, in order. The way in is the interview: email first, under a one-message promise, then a short branching sequence about your estate — lane count, refresh timing, receiving posture — ending in a written read for your situation. The final step locks.
→ Start the interview — it is questions, not a demo.