Case studies
The Box at work
Four scenarios, one per kind of business we build for. Each is an example scenario - the problems are the ones we hear every week, and the system shown is the one our team builds. Results describe the shape of the outcome; real client stories will replace these as clients give permission, with real numbers.
10 Gas Stations, One Phone System
Fuel & Convenience Retail
Example scenario: a family-owned fuel and convenience group - 10 gas stations with stores, a car wash at 3 sites, and one small head office. The owner and one manager run everything.
The problem
Every store had its own phone, its own Facebook page (or none) and its own way of doing things. Calls about fuel prices, diesel, propane, wash hours and fleet accounts rang at whichever store - and rang out after hours. Google listings had wrong hours and old photos at four stores. No one could answer 'which store is growing?' And fleet and dealer inquiries - the high-value leads - arrived by voicemail and were returned a day late.
What our team built
- ✓Foundation: one brand domain and email; every store's Google Business Profile claimed, corrected and synced
- ✓Website: one site with a real page per store - hours, fuel types, wash, directions, its own phone number
- ✓AI Front Desk answering all 10 store numbers 24/7: prices and hours answered, fleet and dealer inquiries qualified (company, vehicles, monthly gallons) and routed to the owner the same hour
- ✓Reviews: automatic requests plus register QR codes, with replies drafted centrally
- ✓Marketing: one social calendar, localized per store
- ✓Custom: an owner dashboard - calls, reviews and inquiries per store - and a one-screen update flow the manager uses once to update every channel
AI in the loop
Answers routine calls (prices, hours, directions), qualifies fleet leads, drafts review replies, and watches listings for changes.
Human in the loop
The owner personally calls every qualified fleet lead. The manager approves price changes and social posts. Any angry caller is transferred to a person immediately.
Results - the shape of the outcome
Illustrative shape of results (replace with a real client's measured numbers before quoting): after-hours calls answered go from none to 100%; fleet inquiries are returned in under an hour instead of the next day; the group's Google rating climbs as every store runs the same review program; the manager's weekly listing and social work shrinks to one short approval sitting.
“Draft placeholder - to be replaced by a real client's words, with permission.”
100+ Sandwich Stores, One Marketing System
Franchise - Quick-Service Food (Multi-Unit Franchisee)
Example scenario: a franchise group operating 100+ sandwich stores across several states under a national brand, supported by a small central team. (The group is described generically: brand names are used only for real clients, with permission.)
The problem
Corporate ran national campaigns; locally, every store was on its own. Store pages and listings were inconsistent. Catering and group orders - the highest-ticket orders - came in by phone during the lunch rush and were missed or written on paper. Reviews piled up unanswered, hundreds a month across the group. Every store manager hired differently. And the owners had no single view of which stores were gaining and which were slipping.
What our team built
- ✓Listings and store pages managed centrally, with one accurate source per store
- ✓Social: one content calendar localized automatically per store - one approval, 100+ localized posts
- ✓Reviews: every review from every store in one inbox, replies drafted in brand voice, problem reviews escalated to a human the same day
- ✓Catering funnel: a catering page per store plus AI phone and chat intake during the rush - menu, headcount, date, delivery - straight into a pipeline
- ✓Hiring funnel: one apply flow per store, with screening questions and interview booking
- ✓Dashboard: rollup by region, district and store - reviews, ratings, leads and catering orders
AI in the loop
Drafts localized posts and review replies, answers catering and FAQ inquiries, screens job applicants, and flags slipping stores in the data.
Human in the loop
The marketing coordinator approves every post batch. District managers own problem reviews. The owners review the rollup monthly with our team. Corporate brand rules are never broken by automation.
Results - the shape of the outcome
Illustrative shape of results: review response goes from a small minority answered to every review answered within 24 hours; catering inquiries during the lunch rush are captured instead of missed; every store posts weekly from a single approval; the coordinator's per-store workload becomes one weekly sitting.
“Draft placeholder - to be replaced by a real client's words, with permission.”
A Medical Practice That Stopped Missing Patients
Healthcare - Multi-Location Medical Practice
Example scenario: a medical practice with multiple clinic locations, one front-desk team stretched across sites, and phones busiest at 8-10 AM and after 5 PM.
The problem
The practice missed calls every day. New patients calling for a first appointment reached voicemail - and simply called the next practice on the list. The front desk spent its day on repeat questions (hours, locations, insurances accepted, directions, forms) instead of the patients in front of them. No-shows ate appointment slots. And reviews mentioned 'can't get through' more than they mentioned care.
What our team built
- ✓AI Front Desk, healthcare-tuned: answers every call 24/7 for every location - books, reschedules and confirms appointments; answers approved logistics questions (locations, hours, insurances accepted, what to bring)
- ✓Reminders and recall: automatic appointment reminders and confirmations; recall campaigns for patients due for follow-up or annual visits
- ✓Intake: digital forms sent before the visit, so the front desk is not re-typing paper
- ✓Reviews: post-visit requests with a private feedback path first, so problems reach the practice before they reach Google
- ✓Reporting: per-location call, booking and no-show dashboard
AI in the loop
Books, reminds, answers logistics questions and recalls patients due for visits - from information the practice approved in writing.
Human in the loop
Staff handle anything clinical - instantly. The AI gives no medical advice and handles no emergencies: emergency calls get the emergency direction and immediate human escalation. The front desk reviews the AI's day every morning, transcripts and exceptions included.
Results - the shape of the outcome
Illustrative shape of results: calls answered go from business-hours-mostly to 100%, around the clock; new-patient voicemail drops to zero - every call is answered or booked; no-shows fall as reminders land; front-desk phone time shrinks to exceptions only; new reviews stop mentioning the phones.
“Draft placeholder - to be replaced by a real client's words, with permission.”
Apartments, Syndication & Crews - One Operating System
Real Estate - Apartments, Property Management & Construction
Example scenario: a real-estate group that owns multiple apartment complexes, raises syndication investments for new deals, runs its own property-management company, and manages several construction and renovation crews across sites.
The problem
Four businesses, four inboxes, no shared system. Leasing leads from listing sites waited hours for a reply - prospects sign with whoever answers first. Maintenance requests arrived by text, voicemail and paper, and got lost. Syndication investors received irregular, hand-made updates that took hours per deal, per month. Construction crews were scheduled by phone and whiteboard, and the owner could not see job progress without driving to a site.
What our team built
- ✓Leasing engine: a website per property plus a group leasing page; an AI leasing assistant answers rent, availability and tour questions 24/7 and books tours; every prospect lands in one CRM pipeline per property
- ✓Resident system: a portal and mobile app - maintenance requests with photos, tracked like tickets (received, assigned, done), with rent reminders and announcements from one screen
- ✓Investor portal: syndications investors see their deals, documents and distributions; monthly updates drafted from live project data
- ✓Crew app: today's jobs, site checklists, photo proof of progress and time logged per job
- ✓Group dashboard: the owner's single screen - leasing, occupancy, maintenance, projects and investor reporting, across every company
AI in the loop
Answers leasing questions and books tours, drafts investor updates from live data, summarizes maintenance and crew activity, and flags stalled tickets and slow properties.
Human in the loop
The leasing team owns every tour and every negotiation. Property managers approve all resident-wide messages. The owner approves every investor update before it sends. AI drafts - humans decide.
Results - the shape of the outcome
Illustrative shape of results: leasing first response goes from hours to under a minute, around the clock; maintenance requests lost goes to zero - every request is a tracked ticket; investor update preparation goes from days of manual work to one review sitting per deal; the owner's daily view goes from phone calls and site visits to one dashboard.
“Draft placeholder - to be replaced by a real client's words, with permission.”
Your business could be the real story here.
Start with the free call. Our team maps where your leads leak and which layers of the Box fix it - you keep the map either way.
Book a free call