Customer loyalty: before vs after with the Masterestaurant method

When the punch card stops being enough?

Your punch card stops working the moment its redemptions come from guests who were already coming, and that number shows up when you cross the redemption log against ninety days of POS tickets.

In the Medellín case that opens this piece, 900 cards issued monthly produced 41 redemptions, roughly 4.5%, and all 41 came from people eating there three times a week; the program was handing margin to guests who needed no incentive at all. The National Restaurant Association reports that 47% of operators run some loyalty scheme, and very few can say how many ADDITIONAL visits it generated. If you cannot separate the incremental visit from the one you had already bought, you are not running a loyalty program: you are running a discount with prettier paper on top of it. Short video shot inside your own kitchen is the most profitable alternative for an independent today because it manufactures intent where none existed, and no stamp card does that.

Reels and TikTok from a real kitchen: the option that builds demand instead of rewarding it

Toast measured in 2024 that 84% of guests would rather see food and drink photos on a restaurant's social feeds than any other content; Tablein records that 67% of Gen Z and 57% of millennials lean on social platforms to decide where to eat, while the National Restaurant Association puts discovery of new dishes there at 74%. WHO IT SUITS: the owner-operator who still works the line and can film ninety seconds between services. Switching cost: a phone, a 25 USD lapel mic, three posts a week held for six months. The hard part was never the gear. It is showing up weekly. A permission-based WhatsApp list is the only loyalty asset you carry with you if you switch POS tomorrow, and that portability wins half the argument by itself. Meta reports open rates between 85% and 95% on that channel, against a 21% hospitality email average; email still pays, mind you, returning 36 USD per dollar invested according to Litmus 2024 and 42.24 USD according to the DMA.

The WhatsApp list: 2,400 numbers beat 2,400 stamps

Stripo also measured birthday coupons redeeming THREE times more often than a standard emailed offer. WHO IT SUITS: neighborhood restaurants with weekly recurrence and one team member able to segment by slow day. Switching cost: nothing in licensing under the free API tier, plus an hour a week for someone to write and measure. Fatigue is the ceiling: two sends a month, with a real offer, or the number goes cold. Every order arriving through a third-party app costs you between 15% and 30% of the ticket and leaves you no phone number, so anyone selling heavily on platforms should buy their own customer back before anything else. Paytronix measured that 70% of consumers would rather order directly from the restaurant than through an intermediary, and Statista puts first-party web or app preference at 67%; demand for the direct channel already exists, what is missing is the bridge.

Winning back the delivery guest: turning commission into a phone number

That bridge is cheap: a printed card inside the bag with a QR offering something tangible —the house dessert, not a vague 5%— in exchange for the number. WHO IT SUITS: operators with more than 30% of sales on platforms. Switching cost: 40 USD per thousand at the printer, plus the discipline of putting one in EVERY bag. Hiring creators from your own city works as an accelerator once your kitchen is already steady, and the 2025 numbers back it: Get Sauce reports close to 8x return and 30% more reservations in the week following a post. Now the fine print, which is where the money leaks. That spike lasts seven or ten days and then falls; if your service breaks under twenty extra covers on a Friday, the creator merely bought you bad reviews at scale. Restroworks measured that 70% of first-time guests never return, so a wave of new faces landing on a shaky operation evaporates anyway.

Local food creators: the expensive shortcut that sometimes runs cheap

WHO IT SUITS: a business with food cost under control and idle capacity midweek. Switching cost: 150 to 600 USD per collaboration across Latin American markets, or a comped meal if your product photographs well. Build the channel you own before the one that makes noise, because doing it backwards means paying twice for the same audience. Diego F. Parra sequences it this way in Masterestaurant audits: week one, card in the bag and number capture; week two, three Reels of the real kitchen; month three, the local creator, once there is somewhere to land the people who show up. Together that package holds under 300 USD monthly for most independents, against a points program with POS licensing that starts around 79 USD a month and then bills you per active customer. Welcome Back reports 4.8x average ROI on loyalty programs and 90% of operators claiming positive returns, a figure that sounds magnificent until you ask which baseline produced it.

The judgment that orders all four: asset first, reach second

There is almost never a baseline. The metric that decides whether your content works is how many new guests walked through the door this month, not your follower count, and that figure lives in the point of sale rather than inside the Instagram app. At the close of every service, for thirty days, ask one question: first time or repeat. A sixty-cover restaurant moving from 12% to 19% new guests, at a 14 USD ticket, adds roughly 1,700 USD of incremental monthly sales; that is the number justifying the work, and no vanity metric replaces it. I got this wrong for years, recommending stamp cards because the owner saw movement within a week. Movement was not business: it was the same guest, paying less, feeling rewarded for doing exactly what he already did. Keep your punch card if the restaurant lives on heavy foot traffic and a low ticket, like an office café or a bakery, where the decision happens in fifteen seconds at the counter and a stamp genuinely pulls forward the sixth purchase of the month.

When NOT to change anything?

Keep it too if you plan to close or sell within twelve months: seeding an audiovisual community takes six to nine months to gain traction, and that work does not transfer easily.

Leave it alone as well if the real problem sits on the plate, because Circana measured in 2025 that 50% of people who stopped dining out would return at lower prices, a sign that perceived value —not loyalty— governs the decision today. Fix food cost and consistency first; a brilliant program over an inconsistent kitchen only speeds up the bad news. This week, cross your redemptions against the register and count how many came from regulars. The stamp card rewards behavior that already existed; short-form video creates demand where no purchase intent was present, and that is the only honest way to grow restaurant sales without cutting prices. A points program leaves the restaurant no asset: change your POS and the history is gone.

The differences that actually move the register

A WhatsApp base of 2,400 opted-in numbers is a channel you control, with 85% to 95% open rates per Meta, far above hospitality email's 21% average. Delivery turned customer loyalty into a question of who owns the data. Every app order costs you 15% to 30% in commission and yields no phone number; converting delivery into a direct channel, with a printed insert in the bag offering something real, recovers 8% to 14% of those orders within three months. A restaurant's sales funnel in 2026 opens with a 12-second video and closes with an occupied table on a Tuesday. Whoever works only the last stretch competes on price; whoever works the first one chooses which guests to keep. Subscription changes what the business sells: pertenencia instead of plates, and cash flow becomes predictable. It is also the alternative most owners abandon by month four, because it demands delivering value every single week even when the kitchen is burning.