Call & Clock
The call hit mid‑morning: registers offline, online orders stuck, and the weekend rush minutes away. Outage windows are measured in revenue per hour, not in abstract SLAs. We gave an ETA and rolled.
On‑Site Triage
We found a clean sky window behind the building, set the dish, and brought a single protected cable inside. Two access points split the floor—front counter and kitchen. Guest Wi‑Fi stayed off. We bound POS tablets first, then printers, then order tablets.
Stabilize & Improve
Once transactions cleared steadily, we tidied cable runs and stress‑tested bulk printing, inventory sync, and receipt email. Everything behaved. We left a short fallback checklist in case the ISP came back “flaky.”
Outcome
By noon, the line moved and the loudest thing in the room was the espresso machine again. A fast, temporary backhaul saved the day and bought time to resolve the root cause without chaos.
Metrics
| Setup time | ~45 minutes (door to POS online) |
| Wi‑Fi zones | 2 (Front counter, Kitchen) |
| Backhaul | Satellite behind building; protected cable run |
| Data used | 14 GB (day) |
| Peak throughput | 70/10 Mbps (down/up) at counter |
| Devices | 10–15 (registers, tablets, printers) |
| Priority | POS and orders first; guest off |