The Zayos guide to a restaurant app people open
Published July 2026 · Updated July 26, 2026
A branded restaurant ordering app is only worth having if regulars install it and reorder from it. Installs come from moments inside your own restaurant, not from an app store. Measure install rate per order, thirty day repeat rate, and taps to reorder. Everything else is decoration.
Why do most restaurant apps get installed once and never opened?
Because the app was treated as a thing to build rather than a habit to earn. A restaurant spends money on the build, announces it once on Instagram, and then waits. Nothing arrives, because an app is not a marketing channel until enough people have it, and nobody wakes up wanting to install a restaurant app.
The build is the easy half. Adoption is the whole job, and it is won in five ordinary moments inside your restaurant rather than in an app store.
It is worth being precise about why you want this at all. The app is not for discovery. It is for the regular, the person who orders from you twice a month and currently does it through a marketplace that charges 15% to 30% base commission and 25% to 35% blended once delivery, marketing and processing fees stack. Moving that one person to your own channel is worth more than a hundred new app installs from people who order once.
Which technology it runs on matters far less than most vendors want you to believe, and we answer that separately on the online ordering app page. This page assumes you have decided to have one, and asks the harder question: how does it get onto a phone and stay there?
Where does an install actually come from?
Five moments, in order of how much intent the customer has when you ask. Run the first one before you spend anything on the rest.
The confirmation screen, right after their first order
The highest intent moment that exists in your whole business. They have just paid, they are pleased with themselves, and they are already standing on your domain in their browser. One line under the order confirmation, with the two tap instruction for the browser they are actually using, converts better than every other prompt on this list combined. If you do only one thing, do this one.
The text that says the order is ready
Order status messages get opened by nearly everyone, which makes them the most valuable real estate you own. Put the link in that message, not in a marketing blast three days later that competes with everything else in the inbox. A ready-for-pickup text with a single line about saving your restaurant to the home screen costs nothing and rides on a message they were going to open anyway.
The bag
A sticker on the bag or a card inside it, with a QR code that goes to your ordering site rather than a generic follow-us page. The moment someone unpacks food at home they have both hands free and thirty idle seconds. It is the only offline touchpoint that reaches delivery customers, who are exactly the people you want off the marketplaces.
The counter and the table
A code by the register and on the table tent, ideally the same one running your QR ordering. Dine in guests are the cheapest audience you will ever have for this, because they are already sitting in your restaurant deciding they like you. Print the code large enough to scan from across a table.
One sentence from your staff
Written down, said the same way by everyone: order from our own site next time, it remembers your order and it keeps more of the money here. Staff scripts beat signage every time because they carry a person behind them. If you want it to actually happen, pay a small spiff per install for the first month and post the count where the team can see it.
What has to happen in the first ten seconds?
An install you earned is lost on the second open, not the first. Five things decide it.
- It opens on the menu. Not a splash screen, not a login wall. Forcing account creation before a first order is the most common self inflicted wound in restaurant apps, and it is invisible because the people it costs you never appear in your numbers.
- Their last order sits at the top with a Reorder button on it. This is the entire product for a regular. Everything else on the screen is for first timers.
- Payment is one sheet: saved card, Apple Pay or Google Pay. If a returning customer has to type a card number, the app has failed at the only job it had.
- Pickup or delivery, and the location, are remembered from last time and shown as a change-able default rather than as a question.
- No forced update prompt, ever. The web based install model means the version they open is always current, which quietly removes the single most annoying thing about restaurant apps.
A regular should reach the Place order button in about three taps and eight seconds. That speed is the entire economic case for having an app, because it is faster than opening a marketplace, and every one of those orders is commission free.
How often can you send a push before people delete you?
The uninstall is silent and permanent. You will never get a complaint, only a number that stops growing. Six rules that keep the channel alive.
Treat push as a budget, not a channel
Roughly one marketing message a week is the practical ceiling for a restaurant. The uninstall is silent and permanent, and you will never know which message caused it.
Order status messages are free
Accepted, being made, ready, out for delivery. These are welcome, unlimited, and they are the reason someone keeps the app installed at all. Do not mix a promotion into them.
Every marketing send needs a reason a customer would recognise
A real special, a closure, a new dish, a holiday menu that opens today. Sending because it is Thursday is how you teach people to swipe you away.
Nothing between 10pm and 9am
Obvious, routinely broken by scheduled campaigns set in the wrong time zone. Check the send window on every automation you turn on.
Segment at the minimum useful level
Ordered in the last 30 days, and lapsed 60 days or more. Two segments beat none by a wide margin, and you do not need more than that in year one.
Keep two other lanes open
Someone who deleted the app may still open a text or an email. Push, SMS and email should carry the same customer record, not three separate lists.
Which numbers tell you the app is working?
Five, and none of them are downloads. Set your own baseline in month one, because published averages blend fifty location chains with single restaurants and tell you nothing about yours.
- Install rate per order
- Installs divided by completed orders in the same week. Track it from week one, because the only benchmark that matters is your own trend. If it does not move after you add the confirmation screen prompt, the prompt is in the wrong place.
- Thirty day repeat rate, installed versus not
- The number that decides whether the app is a channel or a shortcut icon. If people who installed do not order more often than people who did not, stop promoting the install and fix the reorder flow first.
- Taps to reorder
- Count them yourself, on your own phone, as a regular customer would. Open, tap the last order, pay. If it is more than three, do not spend a dollar promoting the app until it is.
- Share of direct orders from installed users
- This is the one that shows up in your bank account. Every order here is an order not paying 15% to 30% base commission, or 25% to 35% blended once delivery, marketing and processing stack.
- Notification opt in and opt out
- Watch the opt out rate weekly. A spike after a send is the message telling you it was not worth sending, and it is the only feedback you will get.
How does Zayos give a restaurant its own ordering app?
Your ordering site on your own domain is also the app. Diners add your icon to their home screen from the browser in two taps, and from then on it opens full screen with their saved cards, order history and remembered modifiers. There is no download wall for a first time customer and no app store review sitting between you and a menu change.
Checkout is Apple Pay, Google Pay or a saved card through Stripe, so a regular reaches Place order in about eight seconds. Re-engagement runs from the customer record you own: web push for opted in regulars, SMS through Twilio with A2P 10DLC registration handled, and email through Resend. Every diner who orders becomes a row you keep, with name, phone, email and order history, and you can export it.
Orders from the app land on the same kitchen tablet as everything else. On Operator plus Marketplace, that tablet also shows Uber Eats, DoorDash and Grubhub tickets ingested through Otter, so the app is the front door of a system rather than a fifth screen for the expediter to watch.
Zayos is $499 per location per month for Operator, $599 for Operator plus Marketplace, and $699 for Concierge. Month to month, no setup fee, no commission. The restaurant keeps 100% of food revenue and tips; the diner pays $0.99 pickup, $2.99 delivery, nothing on dine in, and 10% on catering.
Naya Grill in Pompano Beach and West Palm Beach is live on Zayos and taking orders today, keeping $48k+ a year that used to go to commissions at 3,000 orders a month. Storefronts are built and ordering opens at launch for Shishka Lebanese Grill, La Vie Mediterranean, Yummy Grill, aura, Courtyard Cafe, Mr. Smoke and Yalla Market. Most operators are live in under two weeks. See how it works or the plans.
Restaurant ordering apps, answered
Do I need a native app in the App Store, or is a web based app enough?
How do I get customers to install my restaurant app?
How often should a restaurant send push notifications?
What should I measure to know if the app is working?
Do my customers have to download anything to order?
Is the app included in the Zayos price?
Own the reorder.
The free AI report shows what your regulars are costing you on the marketplaces right now. 60 seconds, no card.