Rolling out a POS update across retail locations

## What gets trained A POS update has three training surfaces: (1) the new transaction flow — how to ring up a sale on the new interface, (2) edge cases — refunds, exchanges, gift cards, layaway, returns without a receipt, and (3) reporting and end-of-day — the X-out and Z-out procedures, the till reconciliation, the manager override flow. The rookie mistake is treating only surface 1 as training. Most cashiers learn the new transaction flow in 20 minutes — it's similar enough to the old one. Where they break is on the edge cases: a customer wants to return a $40 item without a receipt, the new interface has a different override sequence, and the cashier can't process the return. That's where customer complaints come from. Train all three surfaces, not just the first. ## The three-mode rollout The rollout itself should run in three modes. Mode 1 — sandbox training: every cashier completes the training on a sandbox copy of the new POS before launch. Mode 2 — parallel run: for the first 3–5 days after launch, cashiers can fall back to the old interface if they hit a problem they can't resolve. Mode 3 — live only: the old interface is decommissioned and only the new one is available. The parallel-run window is what most chains skip. They cut over hard, and any cashier who hits a problem stalls the queue and pages a manager. With a parallel run, the cashier resolves it on the old interface, learns the gap, and gets a 2-minute follow-up training pushed. ## The training cadence For a typical 60-store retail chain, the training cadence runs: day -10 sandbox training assigned to every cashier, day -3 hard completion gate (anyone not at 100% on the basic flow gets re-assigned to a slow shift for completion), day -1 manager-level POS-update walkthrough (managers run a 30-minute walkthrough at every store), day 0 launch with parallel-run support, day +5 parallel run ends, day +10 final gap-training based on the first week's edge-case patterns. Aristotl handles each of these training pushes — the sandbox training, the day-1 refresher, the day-10 gap training — from the same source POS-documentation. The training stays in sync with the system version because the source doc is the truth. ## Per-store dashboards earn their weight A store manager facing a POS rollout needs to know: which of my cashiers have completed the training, who's at risk on launch day, who needs an extra shift to catch up. The HQ-only dashboard doesn't help them. Per-store dashboards in the franchise-manager scope let each manager run their own readiness check and intervene where needed. ## What HQ tracks Four metrics, daily during the rollout window: training completion rate by store, transaction error rate by store (from the POS itself, not training data), customer-complaint rate by store, and refund rate by store. The first is your training-side metric. The other three are leading indicators that some training didn't land. ## What good looks like A well-run POS rollout has every cashier at 100% basic-flow completion before launch, transaction error rates within 10% of baseline by day +7, and zero stores with significant revenue dips during the rollout window. The 10-day cadence with a parallel-run window is the operational difference between rollouts that go quietly and rollouts that show up in the next earnings call.

Ready to put this into practice?