Restaurant Marketing: Own the Guest, Not the Delivery App

A guest orders from you three times a month. Same dish, same generous tip, reliable as rent. You have never learned her name, her email, or her phone number, because every one of those orders arrived through DoorDash, and DoorDash does not share the guest. It shares the ticket, minus 15 to 30 percent, and keeps the relationship.

Run that math across one slow Tuesday and it looks survivable. Run it across a year of a loyal regular and it is brutal. On a plate carrying a 6 percent net margin, a 25 percent commission does not trim the profit on that order. It erases it and reaches into the next one. You are paying a toll to a company that owns your customer, on a customer who found you first and would happily order direct if you had ever given her a way to.

That is the trap under most restaurant marketing. The delivery apps, the reservation platforms, and the review sites did not just take a slice of your transactions. They took the one asset that compounds: the guest's contact information. This post is about getting it back.

How do restaurants get more customers?

The durable answer is not more marketplace listings. It is to capture the guest's contact information on the first visit and own the channel you reach them through afterward. Restaurants that grow do two things the apps are built to prevent: they take orders on their own site, and they build an email and text list they control, so the second visit costs nothing to trigger.

Almost everything a hungry stranger does before they eat runs across a rented surface. They find you on Google or Yelp, they book on OpenTable, they order on DoorDash. Each handoff feels free because no invoice arrives at the moment of the tap. The invoice arrives later, as commission, as a per-cover fee, as an ad slot you now have to buy just to appear above the competitor two blocks away. And at every handoff, the platform keeps the data and hands you the leftovers.

The fix is not to boycott the marketplaces. They are genuine discovery channels, and a new guest found on DoorDash is still a new guest. The fix is to treat that first marketplace order as a lead, not a relationship, and to move the guest onto rails you own before the second order. Discovery you can rent. The list you cannot afford to.

How much do DoorDash, OpenTable, and Yelp actually cost?

More than the line item. DoorDash and similar apps take roughly 15 to 30 percent of each order, OpenTable charges cover fees on reservations you often sourced yourself, and Yelp sells ads against your own listing. The visible cost is the commission. The hidden cost is that none of them let you keep the guest's contact information, so you re-buy the same customer every time they return.

Compare two restaurants with identical food and identical demand. The first runs everything through the apps. Every order is a fresh rental: the platform found the guest, the platform kept the guest, and the next visit requires the platform again, at the same cut. Volume goes up and margin stays flat, because the toll scales exactly with the traffic. The second restaurant captures a phone number on the first order and texts a return offer directly. The second visit costs the price of an SMS. By the tenth visit the two businesses are not in the same industry anymore.

Every delivery-app order rents a relationship you already paid to earn.

This is the same argument we make about email in the list is the asset: an audience you own is the only marketing channel that gets cheaper the longer you run it, because you already paid the acquisition cost once. A rented relationship never compounds. It resets to full price at every transaction, forever. It is also the same fee math that eats live events alive, which is why we wrote a whole piece on marketing a live event without losing a third of the gate to fees. Tickets, tables, and takeout orders all leak profit through the same hole.

Before you can convert marketplace demand into owned demand, you need to know how findable you are without paying for the privilege. If your Google Business Profile, your map-pack ranking, and your own site are weak, you are handing the apps leverage they should not have. Our free Pre-Flight Check audit reads your local presence, site health, and conversion readiness in a couple of minutes, and the local SEO playbook covers how to rank the map pack on purpose so more of the demand you generate lands on a surface you control.

How do you build a restaurant email and text list you own?

Put a first-party ordering and reservation flow on your own domain, then make contact capture a byproduct of a transaction the guest already wants to complete. An online order, a waitlist signup, a loyalty perk offered at checkout, a QR code on the table that leads to your site instead of a third party. Every one of those is a chance to drop an email or a phone number into a database you own, which turns the second visit into a message you send for free instead of a commission you pay again.

The mechanism is boring and it works. A guest places a direct order through your own checkout. The checkout collects an email and, with consent, a mobile number. That contact lands in a CRM you control, tagged with what they bought and when. Two weeks later, a slow Wednesday needs covers, and you send one text to the four hundred people who ordered in the last ninety days. No commission, no ad spend, no platform standing between you and a customer who already likes your food. That is what owning the rail buys you.

We have not run a restaurant, and we will not pretend we have. But we have built exactly this rail for a business in the same hospitality-commerce category. CineVita, an LA music event, sold its tickets through its own Stripe checkout instead of a marketplace, which meant every ad dollar traced to a named purchaser and the traffic it owned paid zero marketplace fees. Underneath sat a CRM of more than 40,000 contacts that got activated for the launch, plus three landing variants. Swap the word ticket for table and reservation and the architecture is identical: own the checkout, own the list, and stop paying a toll to reach people who already chose you.

The owned restaurant stack, in order

You do not need to rip out the apps this week. You need to build the owned layer alongside them so demand has somewhere to go that you control. Build it in this sequence, because each piece feeds the next.

• First-party ordering on your own domain. Online orders and reservations that run through a checkout you own, not an app that rents you your own guests. This is the collection point for everything downstream.
• A CRM that captures contacts automatically. Every direct order writes an email and, with consent, a phone number into one database, tagged by order history. Not a spreadsheet, not the delivery app's walled garden. Your list, exportable, in your name.
• Owned email and SMS. The channels you can send to for the cost of a message, not a commission. This is where the return visit gets triggered without paying to reach a customer twice.
• Local presence that feeds the top of it. A strong Google Business Profile and map-pack ranking so new discovery lands on your site, where the capture happens, instead of on a marketplace that keeps the data.
• A reason to opt in. A first-order discount, a loyalty tier, an early table for regulars. Treat the perk as the price you pay once for a contact you can then reach for free the rest of the year.

The apps stay in the mix as a discovery channel, which is what they are actually good at. The difference is that a guest who arrives through DoorDash gets converted into a contact you own, so the third and fourth visits belong to you. That is the whole game: rent the introduction if you must, but never rent the relationship.

None of this is exotic technology. It is a checkout, a database, and two messaging channels, wired so they hand you the customer instead of the platform. Most restaurants already pay for three overlapping tools that do slivers of this badly and share none of the data. The work is consolidating them into one owned stack that actually keeps what it collects.

If you are watching the commission line climb every month and treating it as the cost of doing business, that is the reframe worth making: it is not a cost of doing business, it is rent on an asset you should own. We build the ordering, the capture, and the list as infrastructure you keep outright, so the traffic you already earned stops leaking through someone else's checkout. Book a call and we will map which piece of the owned stack pays you back first.