For the VP of store systems at a grocery retailer — the owner of the lane, the December-2027 date, and the largest budget any seat in this programme touches.
There is a date with your name on it: GS1 Sunrise 2027 expects retail point-of-sale to scan and process 2D barcodes and extract the GTIN by the end of December 2027. It is GS1's programme — an industry commitment, not a law — and GS1 US is explicit that the full capability set of 2D is not required by that date. But the baseline it does set lands on hardware and software you own, across every lane in every store, and the arithmetic of getting there is the subject of this brief. The arithmetic is not friendly, and it is not complicated either. It is six quarters long.
Backward from December 31, 2027
Schedule the programme from the deadline toward today and count what each phase honestly takes for a chain measured in hundreds of stores and thousands of lanes:
| Phase | What happens | Honest duration | Must start by |
|---|---|---|---|
| Software cutover | POS software version that parses 2D payloads (element strings and Digital Link URIs) live chainwide; fallback plans retired | 1–2 quarters | Mid-2027 |
| Wave rollout | Scanner replacement/reconfiguration by banner and region; night work, per-store validation | 2–3 quarters | Early 2027 |
| Pilot and certify | One banner, real assortment: prove the lane reads both payload forms, extracts the GTIN, and rings at production speed; certify scanner models and POS builds | 1–2 quarters | Q4 2026 |
| Budget and procure | Capital approval, vendor selection, hardware lead times | 1–2 quarters | This cycle |
Today is July 31, 2026. Between now and the end of December 2027 there are six quarters. Sum the middle of each range above and the programme consumes all six with no slack. Sum the pessimistic edge and you are already late unless a phase compresses. Either way the controlling fact is the same: the first phase is budget, and the budget window that reaches a Q4-2026 pilot is the one open right now. A capital committee that defers this one cycle has not delayed the programme by a quarter — it has moved the pilot into the rollout window and converted a scheduled programme into a scramble.
That is the whole argument for urgency, and notice what it does not rest on: no regulation, no vendor FUD, no forecast of ours. A calendar and a division problem. The companion brief on whether your checkout scanners must be replaced at all covers the hardware-by-hardware reality — many imagers need configuration and software, not replacement — which is exactly the kind of fact a pilot establishes and a deferred budget never learns.
The three cost lines — and the one nobody budgets
Every Sunrise capital request has two visible lines and one invisible one:
- Scanners — the imager population that can physically decode QR and GS1 DataMatrix at lane speed. Partly a replacement problem, partly a configuration problem; the pilot tells you the mix.
- POS software — the version that parses GS1 application identifiers and Digital Link URIs and extracts the GTIN for transaction processing. Scheduled around your POS vendor's release train, which is a calendar you do not control and should get in writing this quarter.
- The record layer — the line item that decides whether the refresh was an upgrade or just an expense. A Sunrise-capable lane reads a richer mark: lot, expiry, serial, a resolvable URI. The transaction log the lane writes to keeps none of it — a lane transaction is not an event record — so a chain that funds only lines one and two arrives in 2028 able to read data it has nowhere to put. The richer read gets discarded at the moment of capture, chainwide, at capital-programme prices.
Line three is adjacent to the hardware budget, not inside it — which is precisely why it gets missed: it belongs to no vendor quote. It is also the cheapest of the three lines, because it is not a per-lane cost at all. The honest framing for the committee: lines one and two make the chain Sunrise-capable; line three makes Sunrise worth having been through.
The cooperative complication
If you operate as a cooperative or multi-banner group, add the governance quarter: the org that decides is not always the org that operates. Corporate certifies scanner models and POS builds; independents own lanes and capital appetites. A programme plan that works in a centralized chain silently loses a quarter to alignment in a cooperative — schedule it as a phase, with named owners per banner, or discover it as a delay.
The one-paragraph brief for your capital committee
Take this paragraph and put your numbers in it:
"GS1 Sunrise 2027 expects our lanes to read and process 2D barcodes and extract the GTIN by end of December 2027. Working backward — software cutover, wave rollout across N stores and M lanes, pilot and certification — the pilot must run by Q4 2026, which requires approval this cycle. The request has three lines: scanner population (mix of replace and reconfigure, pilot-determined), POS software (vendor release-train dependent), and the event-record layer that retains what the upgraded lanes read. Deferring one cycle moves the pilot inside the rollout window and converts scheduled capital into unscheduled risk."
Six quarters is enough — exactly enough, for a programme scheduled now by the person who owns the date. The read for your seat, including how the record layer prices against the lane refresh and what the FSMA-204 receiving clock adds at your other door, starts at get started: address first under the one-message promise, then questions that branch on your chain's shape, ending in a written read that locks when it ends. Every open gap on our side is a named entry in the open P0 ledger — a claim on this origin that outruns that ledger is a defect you should report.