Introduction

Case Study: 100-Location Franchise Unified Tech Stack documents the journey from tool sprawl to a governed, AI-enabled platform. Before consolidation, each location ran its own mix of CRM, phones, and ad tools. Data lived in silos. Duplicates and territory disputes were common. Today, the franchise runs a single source of truth with geo-aware routing, marketplace scheduling, consent management, and BI dashboards that finally agree with POS revenue.

Principle: centralize data and governance; localize service and storytelling. This keeps the brand compliant while letting each city shine.

1) Stack Overview: What We Unified

2) Territory Rules: Zip, Radius, Isochrones

The franchise moved from ad-hoc “we cover everything within 25 miles” to governed maps. Urban areas used drive-time isochrones; rural regions used zip clusters with guardrails to protect crew time.

// Territory Table (excerpt)
territory_id,name,method,zips,max_drive_min,phone,lp_url,utm
CO-DEN-N,Denver North,isochrone,"802xx, 800xx",35,(303) 555-0110,/denver-north,FR_CO_DEN_N
CO-DEN-S,Denver South,isochrone,"801xx, 802xx",35,(303) 555-0120,/denver-south,FR_CO_DEN_S

3) Lead Routing Logic & SLAs

// zip/GPS → territory → queue with SLA
lead = capture()                           // form/chat/call/marketplace
geo  = geocode(lead.zip || lead.latlon)
terr = lookupTerritory(geo)
queue = terr.capacity_ok ? terr.queue : terr.overflow_queue
assign(lead, queue)
SLA: AI_first_reply <= 20s; Human_handoff <= 5m; Book_offer same_day

Every assignment gets an audit_id so disputes are resolved with data.

4) AI Responders: From “Hi” to Booked

• Greets in seconds with location-aware copy.
• Checks inventory/eligibility; offers next timeslot.
• Writes notes back to CRM, tags territory and source.
• Escalates to human if risk keywords or complex requests appear.

// AI handoff guardrails
if lead.intent in ["emergency","legal","custom_quote"]:
  escalate_to_human(now)
else:
  propose_timeslots(geo, service, capacity)

5) Marketplaces & Listings at Scale

• Territory accounts with scheduled posting windows.
• SKU/asset registry to avoid photo and copy collisions.
• Negative geos on paid campaigns to prevent cannibalization.
• Minimal overlays; policy-safe captions with disclosures.

6) CRM ⇄ POS: Source-to-Revenue Matchback

Daily ETL syncs orders to the warehouse; BI matches revenue to UTM and territory. Finance and marketing finally share the same numbers.

// ETL mapping (simplified)
pos.order_id → dw.order_id
pos.territory_code → dw.territory_id
pos.amount → dw.revenue
crm.lead_id → dw.lead_id
crm.utm_campaign → dw.utm_campaign

7) Phones & Call Tracking: Territory Numbers

• Unique numbers per territory and per campaign family.
• IVR respects hours and language; emergencies route to nearest crew.
• Call outcomes posted back to CRM with recording links.

8) Consent, Identity, and SSO

• SSO with role-based access (RBAC) for franchisees and staff.
• Consent ledger stores channel preferences and timestamps.
• APIs enforce read/write of consent fields across tools.

9) Data Warehouse & BI Dashboards

10) Local Landing Pages & UTM Standards

• Per-territory numbers and forms; above-the-fold CTA.
• UTM pattern: utm_source, utm_medium, utm_campaign=terr_{id}
• Localized testimonials, hours, and service radius.

11) Brand Templates with Local Fields

{City/Neighborhood} • {Service} • {Earliest Availability}
Reply “BOOK” for times or “QUOTE” for pricing. License {#}, Territory {ID}

Keep overlays light; keep logos small; disclose licenses consistently.

12) Governance: Guardrails, Disputes, Audits

• Tooling council approves apps and vendors.
• Dispute SLA 24–48h; nearest-crew-wins tiebreaker for field work.
• Audit logs for all reassignments and territory edits.

13) Capacity, Scheduling, and No-Show Saves

• Cluster jobs by drive-time; show travel on calendar.
• Auto-reminders and self-reschedule links lift show rates.
• No-show win-back: 3-touch sequence with new time blocks.

14) KPIs: What We Measured Weekly

Use consistent time zones and week definitions for apples-to-apples comparisons.

15) Security & Access: Roles and Least Privilege

• Role scopes: Franchisee, Manager, Agent, Analyst, Admin.
• PII minimization and masked fields for exports.
• Quarterly access review and offboarding automation.

16) 30–60–90 Rollout Plan (Pilot → Scale)

Days 1–30 (Pilot Foundation)

1. Document systems and owners; ship the territory table.
2. Integrate CRM, phones, and AI for 5–10 locations.
3. Define SLAs; launch BI prototype; capture baseline.

Days 31–60 (Momentum)

1. Expand to 25–40 locations; enforce negatives and SKU registry.
2. POS→warehouse ETL live; attribution reports match finance.
3. Train managers; certify on disputes and audits.

Days 61–90 (Scale)

1. Roll to 100 locations; automate overflow and surge playbooks.
2. Quarterly business reviews; rationalize tool portfolio.
3. Publish change logs and roadmap; set next-quarter tests.

17) Playbooks: Emergency, Surge, Off-Season

18) Results: What Changed in 90 Days

• Reply times fell from minutes to seconds (AI) and <5m (human).
• Duplicate outreach dropped—guardrails + audit trails.
• BI revenue matched finance; territory dashboards drove staffing decisions.
• Marketplace compliance issues decreased with standardized templates.

19) Troubleshooting Matrix