Retail & Hospitality Technology

How It Works & Technology Overview

Retail and hospitality technology is the trade of repeatability: POS lanes, drive-thru systems, kiosks, menu boards, and back-office infrastructure deployed across fleets of sites. The economics live in the pilot-to-Nth-site curve — the pilot (first 2–3 sites) is funded engineering that produces a runbook (per-site execution document: photos, cable counts, mount specifics, acceptance checklist) and a kitting spec (every site's materials arriving as one labeled, pre-configured kit); then waves execute at a fraction of pilot labor with night crews and a remote command center handling activation and escalation. The governing constraint: the business cannot close — POS cutovers run between close and open against a fixed window, with tested rollback (old system selling by morning) as part of every plan. Site architecture is consistent: lane/counter devices (terminals, payment devices, printers, scanners) → back-office switch/router in a closet that's really a shelf → WAN (broadband + LTE backup) → cloud POS or franchisor data center. Drive-thru adds timers, headsets, vehicle-detection loops, and outdoor menu boards — with civil-work adjacency (conduit under lanes, bollards, canopy mounts) that drags in site-work trades.

Common Deployment Scenarios

• POS refresh waves: fleet-wide lane hardware/software replacement; night cutovers, per-site 4–8 hour executions, rollback-ready.
• New store / new franchise openings: full technology stack on the GC's construction schedule; the opening date does not move for us.
• Drive-thru modernization: timer systems, wireless headsets, loop replacement, digital menu boards; 1–2 nights per site plus civil coordination.
• Kiosk/self-checkout rollouts: floor-anchored units with power/data below, ADA-compliant placement, and content/software owned by others — physical + network scope for us.
• Connectivity refresh: SD-WAN/broadband + LTE-backup standardization across a fleet; circuit lead times (30–90 days) are the schedule, not the install labor.

Products, Vendors & Characteristics

• POS platforms: Toast, Square, NCR/Aloha, Oracle Micros/Simphony, Par (QSR), Lightspeed — usually chosen by the brand/franchisor before we arrive; our scope is physical deployment, networking, and cutover, not platform selection.
• Payment devices: Verifone/Ingenico terminals with P2PE key-injection chains — payment hardware often ships pre-injected on a controlled chain of custody; never improvise substitutions.
• Drive-thru: HME and Panasonic dominate timers/headsets; loop detection needs saw-cut or preformed loops (civil adjacency); digital menu boards follow the digital-signage discipline (high-brightness for outdoor, content ops owned by marketing).
• Networking: Meraki/Fortinet-class managed SD-WAN with LTE failover is the fleet default — remote manageability is the product; local IT does not exist at site.
• Franchise vs. corporate matters commercially: corporate programs sign once and schedule centrally; franchise programs need per-owner approval, scheduling, and sometimes payment — coordination labor that scales with owner count, not site count.

What Makes It Hard (Gotchas & Failure Modes)

• The identical-sites myth: even franchised builds drift (remodels, local code, landlord quirks, that one store with the closet in the basement) — an early survey wave converts wave pricing from guess to commitment; every unsurveyed variance is unpriced night work.
• Night-work access failures: alarm codes that don't work, no-show key holders, silent alarm rolls to police — access confirmation procedures per site, or crews burn nights and reschedule fees stack.
• Cutover without tested rollback: if lanes aren't selling at open, the cost isn't labor — it's the client's revenue and the program's reputation.
• Kitting skipped as a savings: midnight on-site assembly multiplies per-site hours and error rates; the kit is the cheapest labor ever bought.
• Circuit lead times discovered late: 30–90 day provisioning against an already-scheduled wave collapses the calendar.
• Scope creep at the door: store managers asking night crews for "one more camera" — the runbook is the scope; changes route through the program manager or fifty favors become an unbilled site.
• Payment-device chain of custody: P2PE devices can't be casually swapped or sourced; a broken chain is a compliance event, not a hardware problem.

Effort & Estimation Patterns

• Pilot sites: 2–3× target per-site labor, plus engineering time for the runbook and kitting spec — the tooling investment the program amortizes.
• Nth-site execution: two-tech night crew lands a standard POS-lane or small-format refresh in 4–8 hours per site with kit pre-staged; drive-thru systems: 1–2 nights per site depending on civil scope.
• Site surveys: 1–2 hours per site plus travel, run as an early wave and priced explicitly.
• Travel/mobilization: route waves geographically; per-site travel is a visible cost line; out-of-region waves carry lodging/per-diem math.
• Remote command center: staffed hours per cutover night (activation, testing, escalation) — a program cost line, not absorbed overhead.
• Per-site documentation + photo acceptance: 30–45 minutes per site, non-negotiable — it's what makes site 30's future service call answerable without a truck roll.
• Revisit contingency: 5–10% of sites need a second touch — carried in program price, not discovered in margin.
• Recurring: fleet support (monitoring, RMA logistics, per-site service SLAs) is the natural annuity — a 50-site fleet generates service volume from day one, and whoever holds the documentation wins that contract.