Should Restaurants Offer Delivery? An Honest Guide

Delivery can feel like found money: the same kitchen, the same menu, a wider net of hungry people. But the question owners should actually ask is not whether delivery is popular. It is whether delivery is profitable and on-brand for your specific food. So, should restaurants offer delivery? Sometimes yes, sometimes no, and the answer lives in the numbers and in how your dishes behave inside a paper bag on a twenty-minute drive.

This is a channel with its own economics, not a free extension of your dining room. Treat it that way and it can add real revenue. Treat it as a switch you flip, and it can quietly erode the margin you already fought to protect.

Delivery is a channel, not free money

Every order that leaves your door carries costs the dine-in version never had: a container, a bag, a label, someone to pack it, and either a commission to a marketplace or a driver you pay yourself. None of that shows up when you look at menu price alone.

The honest framing is simple. Dine-in and delivery are two different businesses that happen to share a stove. One seats guests, sells drinks, and turns tables. The other ships food and lives or dies on packaging, timing, and fees. A dish that prints money at the table can lose money on the road once you account for everything the road adds.

Which foods travel, and which do not

Before any spreadsheet, look at your menu with cold eyes. Food that travels well holds heat, holds texture, and forgives fifteen to thirty minutes in transit.

Travels well:

• Braises, stews, curries, and saucy dishes that stay hot and only improve as they sit
• Pizza, sandwiches, wraps, and burritos built to be handheld
• Fried items that are par-cooked and finished to order, packed with a vent
• Grain bowls and salads where the dressing ships on the side

Travels badly:

• Anything crisp that must stay crisp against steam: thin fries, tempura, delicate fried fish
• Plated dishes whose value is the presentation on the plate
• Soufflés, medium-rare proteins, ice cream, and anything on a precise temperature knife-edge
• Composed dishes where sauce, protein, and garnish must arrive separate and don't

This single filter can decide the whole question. If your signature dishes arrive soggy, cold, or collapsed, delivery will not just underperform. It will actively teach new customers that your food is worse than it is. That reputational cost is real and rarely shows up in the sales report.

The three models and their trade-offs

There are three broad ways to run delivery, and each trades money for effort differently.

Third-party marketplace. You list on a large app, they bring demand and drivers, and they take a commission, widely reported in the 15–30% range depending on market and plan. Lowest operational lift, highest cost per order, and you rarely own the customer relationship or the data behind it.

Your own site plus your own couriers. You take orders on a channel you control and either hire drivers or use a per-delivery courier service. More work and more coordination, but you keep the margin, the customer, and the brand experience end to end. Running your own commission-free online ordering is what makes this model add up, because the fee that would have gone to a marketplace stays in your business.

Hybrid. Many restaurants list on a marketplace for reach and discovery while steadily pushing repeat customers toward their own channel with inserts, better pricing, and loyalty. You accept marketplace economics on new customers and protect margin on the ones who come back.

The marketplace math

Here is where good intentions meet arithmetic. Restaurant margins are thin to begin with, and a percentage taken off the top of an already-thin margin can erase the profit entirely. This is worth modeling before you commit, and it is the core of the real cost of marketplace commissions.

Consider a hypothetical order. The numbers below are illustrative, not a benchmark.

• Menu price of the order: $30.00
• Food cost at 30%: –$9.00
• Packaging (containers, bag, label): –$1.50
• Marketplace commission at 25%: –$7.50
• Labor to prep and pack, allocated: –$6.00

Add those costs: $24.00. Contribution before overhead: $6.00, or about 20% of the ticket. Now recall that rent, utilities, insurance, and management still come out of that $6.00. On a slower day, or at a 30% commission, or with a dish that needs more packaging, the same order slides toward break-even or a loss.

Run the same order on your own channel and the $7.50 commission line largely disappears, replaced by a smaller courier or card fee. That single change is often the difference between a channel that funds your business and one that funds someone else's.

Packaging: cost and quality both matter

Packaging is where owners tend to guess. Good containers cost more, and they are usually worth it, because they protect the one thing delivery puts at risk: the quality of the food on arrival. Vented lids for fried items, leak-proof bowls for saucy dishes, and separate cups for anything that should stay crisp or cold all cost real cents per order. Budget them explicitly, per dish, and put the number in your unit math rather than treating packaging as a rounding error.

The hidden load on the kitchen

Delivery lands on a line that is often already busy with dine-in. Tickets arrive on a second screen, on someone else's timing, during your rush. Staff stop to pack, chase couriers, and field "where is my order" questions. Throughput for your seated guests can suffer at exactly the wrong moment.

This operational drag is the cost owners most often forget. Before adding delivery, ask honestly whether your kitchen has slack during peak hours, or whether every new channel steals attention from the guests already in the room.

The case for delivery, done right

None of this is an argument against delivery. Done well, it adds incremental revenue from customers who would never have visited, extends your reach beyond walking distance, and fills the slow mid-afternoon and early-evening stretches when your dining room is quiet and your fixed costs run anyway.

The condition is protecting margin, and the clearest way to do that is to take as many orders as you can on channels you own. Use marketplaces for what they are genuinely good at, discovery, and work to convert those diners into repeat customers who order direct.

Start this week

You do not need to decide the whole strategy at once. This week, do three concrete things. First, pick the five dishes you would actually put on delivery and be ruthless about which ones travel. Second, build the unit-economics math for each, packaging and fees included, so you know the real margin per order rather than the top-line sale. Third, run a small, time-boxed pilot, perhaps a few off-peak hours on your own channel, and measure profit after every cost, not just orders placed. If the pilot clears real margin and your food arrives the way you'd serve it, expand deliberately. If it doesn't, you will have learned that for the price of a slow Tuesday instead of a slow year.