Background
Imagine a three-unit pizza concept in the Phoenix metro — one flagship in a dense walkable corridor, one in a strip-mall position near a school, one in a newer mixed-use development that hadn’t quite hit its stride. New York-style pies, a respectable wing program, a small craft-beer list. Dinner was strong. Friday and Saturday nights routinely hit ticket times of 28 minutes and a packed dining room.
Lunch was a different story. The flagship held its own because of foot traffic. The two satellites bled money five days a week, with covers running 30 to 40 percent below what the prep schedule was built for. The operator had been comping pizzas to ad-hoc Facebook giveaways just to put bodies in chairs and stop the dining room from looking dead from the street.
The problem
Two compounding issues showed up the moment we pulled the data.
First, there was no loyalty list to speak of. The POS captured names and phone numbers on online orders, but nobody had ever exported them. There were roughly 11,000 unique phone numbers across the three stores’ order history — and zero of them had ever received a marketing message. A goldmine sitting untouched.
Second, the brand had no daypart-specific identity. To the lunch customer, the restaurants looked like a dinner spot that happened to be open at noon. There was no lunch combo, no lunch-specific landing page, no reason to choose Tuesday at 12:30 over any other quick-service option in a six-block radius.
The operator’s instinct had been to discount harder — a third Groupon run was being scoped when we started the snapshot. That would have papered over the symptom and made the underlying margin problem worse.
What we shipped
The snapshot focused on two big pieces, both shipped inside the first three weeks.
The snapshot loyalty list. We pulled 11,000 order-history phone numbers into GHL, ran them through a compliance check, and built a re-permission flow: a friendly SMS that introduced the new loyalty program, offered a free order of garlic knots on the next visit, and asked recipients to opt in. The opt-in rate landed around 38 percent over the first two weeks, giving the operator a 4,200-name SMS list seeded entirely from existing customers. No paid ads, no list rental, no Instagram contest.
The Tuesday family pack. Every Tuesday morning at 10:45, the full loyalty list received an SMS: a family pack — large pizza, half-dozen wings, breadsticks, two-liter — for $46, dine-in or pickup, Tuesday only. The price was deliberately set above any prior promotional discount the brand had run, because the goal wasn’t volume at any cost; it was a profitable daypart. Order-ahead links inside the SMS went straight to the brand’s existing online ordering, with a Tuesday-only menu module flipped on automatically.
A handful of smaller pieces rounded it out:
- A birthday-month coupon (one free large pizza, valid the diner’s birthday month, dine-in only — to drive an in-restaurant visit rather than a cheap pickup).
- A win-back flow for any number that hadn’t ordered in 60 days, with a softer ask: a free appetizer, no minimum spend.
- A simple review-request SMS that fired 90 minutes after a confirmed dine-in visit at the flagship.
30-day outcome
The first Tuesday after launch was the test. The two satellite locations had been averaging 41 lunch covers on Tuesdays. The first Tuesday post-launch, they pulled 71. The flagship, which hadn’t really had a lunch problem, still added a meaningful bump.
By the end of month one, Tuesday lunch covers across the group were up roughly 38 percent versus the prior 90-day baseline. The family-pack average order value sat at $46 — exactly the offer price, because almost everyone took the offer as priced rather than adding to it, which was fine. Food cost on the family pack penciled at 28 percent versus 35 percent on the prior ad-hoc discounting, because the offer was built around margin rather than against it.
The operator also got something less obvious: real visibility into who their customers actually were. The loyalty data showed that 19 percent of names had ordered at least twice in the prior six months — a hidden core of repeat diners the team hadn’t known existed.
90-day outcome
Three things compounded by month three.
Repeat-visit rate climbed from 22 percent to 31 percent. The Tuesday cadence trained guests to expect a reason to come in early in the week. The team layered a “Thursday wing night” offer at half the marketing weight (just the existing loyalty list, no separate ads) and saw a smaller but real Thursday bump.
Review velocity at the flagship roughly tripled. The 90-minute review SMS was doing the work. The two satellites had a harder time because the in-restaurant experience wasn’t optimized for review capture yet — different problem, separate fix.
The win-back flow brought back 340 lapsed diners. Lapsed defined as 60+ days inactive. That’s roughly $15,000 in incremental revenue against zero ad spend, just by texting people who had already bought before.
What’s next
The operator is testing two extensions for the next quarter:
- Daypart-specific menus baked into the loyalty cadence — a breakfast pie at the flagship on weekends, a build-your-own slice combo at the school-adjacent satellite, a happy-hour by-the-slice push at the mixed-use unit.
- Catering capture from the existing loyalty list — about 12 percent of the opt-ins are tied to numbers that have ordered party-size quantities in the past, and the team is building a simple “order catering 48 hours ahead” flow targeted just at those numbers.
Your POS already knows who your regulars are
If you've got an order history sitting in your POS that nobody has marketed to, the snapshot loyalty pull pays for itself before the first Tuesday.
“Two of our three stores ran a 35 percent food cost at lunch because nobody was coming in. The snapshot didn't invent demand — it surfaced the guests we already had and gave them a reason to choose Tuesday.”