Restaurant database funnel: the 2026 numbers and the decision each one should trigger

How many of your guests do you actually own in a database?

Fewer than 10% in most restaurants we review, and that single number explains why social marketing never shows up in the till. A 180-seat room turning twice a day serves roughly 60,000 meals a year;

if your list holds 340 emails, you OWN the contact details of 0,6% of the people who already walked in and paid. Everything else is rented reach. And rent went up: per Marqii, customer acquisition cost climbed 222% over the eight years through 2025, so every guest who arrives, eats, pays and leaves without an identified trace forces you to buy them again at 2025 prices. The decision this figure triggers is straightforward: before approving another peso of paid media, set your monthly capture target as a percentage of covers served, never as a raw follower count. With an average open rate of 25,1% in 2023 according to Omnisend's email, SMS and push marketing report, email reaches one in four subscribers without paying anyone for the privilege.

Email is still the cheapest channel that comes with permission

Compare that against the organic reach of a social post, which no restaurant controls and which shifts the moment the platform decides to shift it. On a list of 4.000 identified contacts, that open rate means roughly 1.000 people reading your Tuesday message; buying 1.000 equivalent reads through paid media costs you every single time, and next time you pay again. I got this wrong for years: I treated email as the old channel next to video, until I put the cost per confirmed reservation of both side by side in the same month. The move is to send on a steady cadence —every fortnight as the floor— because a list that hears nothing for eight months goes cold until the open rate stops being worth what it cost to build. Eight out of ten restaurants were projected to run an active loyalty programme by the close of 2025 according to LoyaltyPass restaurant loyalty statistics, so having one sets you apart from nobody: lacking one leaves you outside.

Loyalty is no longer a differentiator, it is the price of entry

What does set you apart is enrolment pace. Top QSR operators sign up around 110 new members per store each month per Paytronix's Annual Loyalty Report 2024, and that number works as an operating yardstick rather than a report curiosity: 110 monthly sign-ups in a store serving 5.000 covers a month is 2,2% capture, while twelve months at that pace build 1.320 contacts carrying permission, purchase history and measurable frequency. If your programme enrols twenty a month, the fault is not the promotion but the capture point: nobody asks for the detail at the moment the guest is happy, which is when the check lands and not when they come through the door. A restaurant saves an average of 3.600 US dollars a year with QR menus according to the 2025 QR code statistics for restaurant usage published by QR Code, and that printing saving is the least interesting part of the story.

The QR code is not the menu, it is the cheapest capture point you have

What matters is that every scan is an identifiable device inside your room, at a table, at a specific hour, with a spend attached. That said, I keep repeating what people ask me every week: the PHYSICAL menu stays. QR complements it for delivery, accessibility, price changes and analytics on what gets viewed; the printed menu governs service pace, dish narrative and suggestive selling. One operator pulled the printed menu to force scanning and lost average check before gaining a single email address. Use the QR to ask for the contact after dessert, tied to a concrete incentive, and leave the menu exactly where it was. Some 37% of adults order delivery at least once a week and more than 40% order delivery or takeout three to five times a month, both figures from UpMenu's 2024 food delivery statistics report. Translate that into your operation: the slice of revenue riding on a motorbike carries the highest frequency in your whole business and, if it travels through a platform, those people's contact details are NOT yours.

Delivery: high frequency, someone else's data

You cook, you pack, you absorb the commission and you finance a third party's database that later charges you to speak to your own customer. That is where the funnel breaks without anyone noticing, because sales rise while ownership of the guest falls. The decision: drop a printed insert into every package with a repurchase incentive redeemable only on direct orders, then measure how many orders migrate from platform to owned channel each month. A five-point lift in customer retention raises profits between 25% and 95%, according to Frederick Reichheld, creator of the Net Promoter Score and emeritus partner at Bain & Company, and that arithmetic bites with particular harshness in a low-ticket, high-frequency business like ours. The reason sits in the till: the second order carries no acquisition cost, so its contribution margin drops whole to the profit line while the first is still repaying the media that brought it in.

Why the second order is worth more than the first?

With acquisition running 222% more expensive than eight years ago per Marqii, a restaurant that only counts first visits is buying revenue that leaves no profit behind.

Diego F. Parra and the Masterestaurant framework measure the database capture funnel with a single yardstick —the percentage of identified guests— precisely because it is the one marketing metric that survives an algorithm change. Taken together, these three numbers order the budget: capture first, reactivate second, advertise only after that. Say it goes down on a Thursday and your list holds 340 unsegmented contacts: on Friday you have no way to tell anyone that you are open, that the menu changed or that tables are free; by Saturday the till drops and your only exit is buying reach again on another platform, at an acquisition cost 222% higher than in 2017 per Marqii. Hold 6.000 permissioned emails with the 25,1% open rate Omnisend reports, and Friday reaches 1.500 people for free while the weekend holds.

What happens if your Instagram account vanishes tomorrow?

That is the gap between buying attention and buying permission:

attention dies with the campaign, permission sits on your server and works through the following thirty-six months, which is the real horizon over which a restaurant recovers what it spent on marketing. The Bogotá case that opens this piece —412.000 Reel plays, 9.100 new followers, sales down 6%— is precisely what happens when the entire asset is rented. First: 25,1% email open rate (Omnisend, 2023). Action: write every fortnight and purge anyone who has ignored six consecutive sends, because a dirty list drags down deliverability for the clean one. Second: 110 loyalty sign-ups per store per month among the best operators (Paytronix, Annual Loyalty Report 2024). Action: hand that target to your floor manager, split it by shift and review it on Mondays alongside the sales report, not in the marketing meeting. Third: 222% growth in acquisition cost over eight years (Marqii, 2025).

The 3 numbers worth tattooing on your forearm

Action: cap monthly paid media and redirect every extra peso into reactivating contacts you already hold, which cost zero to acquire. And one figure to close on rather than a recap: with the 9,8% rise in menu prices ACODRES reported in 2025 to sustain 98.000 jobs across Colombia, guests are more price-sensitive than ever, and the ones who flinch least are the ones you already know by name. The traditional route buys attention; the MASTERESTAURANT method buys permission. Attention expires when you switch the campaign off, while permission sits on your server working for free across the next thirty-six months, which happens to be the horizon over which a restaurant recovers its marketing investment. The traditional route celebrates the first order; we measure the second. According to Frederick Reichheld, creator of the Net Promoter Score and emeritus partner at Bain & Company, a five-point lift in customer retention raises profits by 25% to 95%, and that arithmetic bites hardest in a low-ticket, high-frequency business like ours.

Five differences that move the till

Where the agency sees a three-step funnel, the operation runs seven: media, profile, message, booking, table, record, repeat visit. The four middle steps belong to the dining room rather than to marketing, and that is precisely where 80% of the acquisition you paid for leaks away. Traditional customer acquisition cost is measured per campaign; ours is measured per record and amortised against accumulated contribution margin from visits two, three and four. A USD 3.40 record producing two USD 42 visits at 58% margin returns USD 48.72: fourteen times the outlay. Online reputation in the traditional model stops at replying to reviews. In ours every reply opens a private channel, and 6.8% of reviewers approached with a personal invitation hand over their WhatsApp, which turns the complaints desk into the third most productive capture point in the business.