Case studies

Real restaurants. Real numbers.

Three operators, three different stacks they replaced, three different outcomes. The before/after numbers are what they tracked themselves — we just compile them.

Tortilla Sol Cantina

Austin, TX · Single-location Tex-Mex · 90 days in

Outcome. Direct order share grew from 22% to 71% of online volume in 90 days, with no extra ad spend.

  • +49pp
    direct order share lift
  • $6.2k
    monthly savings vs. delivery commissions
  • 38%
    repeat customer rate at 60 days
“Before RestaurantGPT, every direct order felt like a small miracle. Now they're the default.”
— Lucia Marrero, Owner

Story

Tortilla Sol was running on a four-year-old Squarespace site, an off-the-shelf online ordering plugin that charged 8% per order, and DoorDash/Uber doing 78% of their online volume at 30% commission. Lucia knew the math was bad — she didn't know how bad until we ran the audit.

In the first 30 days, we shipped a new mobile-first site (built from her Google reviews and existing menu in under 90 minutes), direct ordering with integrated payments and same-day delivery dispatch, and the 24/7 receptionist on her main line. Marketing automation came on in week 3 once the welcome flow was tuned.

By day 60 the win-back flow had pulled 142 dormant customers back, the welcome flow was converting 41% of first-time online orderers into a second order within 30 days, and the receptionist had captured 84 calls outside business hours that previously went to voicemail.

By day 90, direct orders had nearly tripled and the third-party delivery share had shrunk to under a third. Net of platform costs, Tortilla Sol is keeping an additional $6,200/mo in margin.

Foxbridge Coffee & Bagels

Portland, OR · Two locations · 6 months in

Outcome. Consolidated four vendors into one stack; cut monthly tooling spend by $720 while doubling repeat-customer rate.

  • 4 → 1
    vendor stack consolidation
  • 2.1x
    repeat rate at 90 days
  • $8.6k
    annual tooling savings
“We finally have one dashboard for both shops. I can see at a glance which location needs help, which is having a great week, and who our regulars are.”
— Arman Joshi, Co-owner

Story

Foxbridge was running on a stack of four separate vendors: a website builder, an online ordering platform, a CRM tool for SMS marketing, and a separate analytics dashboard. Each charged $80–280/month and barely talked to each other. Arman and his co-owner spent ~6 hours a week stitching reports together.

We migrated their menu, customer list, and brand assets in a single weekend. The new site was live in 90 minutes; the existing online ordering routed straight into their Toast POS so the staff workflow didn't change. Customer data unified into a single profile in the dashboard.

The unlock wasn't cost — it was visibility. Within 30 days they could see that the SE Portland location had a stronger pickup mix and weaker repeat rate than the East Side, and they shifted their email cadence accordingly. By month 6, their repeat-customer rate had gone from 19% to 41%.

Osteria Velluto

Brooklyn, NY · Single-location Italian · 4 months in

Outcome. Captured 312 reservations from previously-missed calls in 120 days, generating an estimated $58k in incremental revenue.

  • 312
    previously-missed calls converted
  • $58k
    estimated incremental revenue
  • < 30s
    average call answer time
“The receptionist alone saved us. We were losing reservations to voicemail every dinner shift. Now nothing slips through.”
— Marco Riccardi, Chef-owner

Story

Marco runs the kitchen at Osteria Velluto nightly. There's no front-of-house manager during prep, which meant phone calls between 3pm and 5pm — peak reservation-booking time — went to voicemail. He estimated they were losing 4-6 reservations a day.

The 24/7 receptionist took over the line. Trained on the menu, hours, dietary restrictions, and reservation rules, it answers in English or Italian, books straight into their Resy account, and texts Marco a summary of every call.

In four months, the receptionist handled 1,840 calls. 312 of them were reservation bookings outside normal staffed hours — almost all would have gone to voicemail. At their average party size and check, that's ~$58k in revenue that previously bled away.

Marco didn't need the website, the ordering, or the marketing automation for the math to work. The receptionist by itself paid for the platform 22x over in the first quarter.