Should Your Restaurant Hire Its Own Delivery Driver?

Your own delivery driver becomes cheaper only when enough orders arrive during the hours you pay for—and your driver can actually deliver them. Compare the full shift cost with the app fees you would avoid. Then check how many customers would still order directly. A lower cost per order does not help if you lose too many orders.

In the example below, a restaurant needs 27 direct delivery orders in four hours to beat a 25% app fee on cost. But if each delivery takes 20 minutes out and back, one driver can complete only 12. The spreadsheet says 27; the street says 12. That is the problem to solve before hiring.

This is a U.S. planning example, checked September 12, 2026. Every dollar amount below is fictional. Use your own payroll costs, contracts, routes and order records.

First, separate ordering from delivery

“We offer our own delivery” can mean three different things:

• Marketplace order, marketplace courier: the customer finds you in an app, orders there and receives delivery through its network. Your restaurant pays the charges in its agreement.
• Direct order, outside courier: the customer orders from your website or by phone. You pay for ordering/payment services and dispatch a courier. You still need to attract the customer.
• Direct order, your employee: you take the order and schedule a driver. You pay for the employee's paid time, vehicle arrangements and other delivery expenses.

There is also a hybrid: customers order in a marketplace and your employee delivers. That can still carry a platform fee. For example, Uber's U.S. pricing page lists Self-delivery separately from marketplace courier delivery and Uber Direct. Check your market and agreement. Hiring a driver does not automatically remove every app charge.

Gather these numbers for one delivery window

Start with a window you might actually staff, such as Friday from 5 to 9 p.m. Don't divide a Friday driver's cost by the entire week's orders.

1. Orders and food sales: count completed delivery orders in that window. Divide their food sales after restaurant-funded discounts by their order count to get an average order. Keep sales tax and employee tips out of food sales.
2. Marketplace cost: read an actual payout statement alongside your agreement. Include commission, separately charged processing, restaurant-funded promotions, fixed fees and expected refund costs. Use the correct fee basis; it may differ from your sales figure. Subtract a discount only once: if it already reduced food sales, do not subtract it again as a cost.
3. Paid driver cost: hours multiplied by the full employer cost per hour. That includes wages and the employer costs that apply, such as payroll taxes, workers' compensation and benefits. Add overtime or other required premiums. Include paid waiting, return travel and closing tasks.
4. Direct-order costs: payment processing, ordering software, dispatch, vehicle cost or reimbursement, insurance changes, marketing and support. Separate costs paid per order from costs paid per shift or month.
5. Travel and demand: where orders go, when they arrive, and how long the whole delivery cycle takes. Count the return to the restaurant.

If you have not opened yet, use quotes and a conservative delivery forecast. A hopeful forecast is not evidence that you can keep a driver busy. Use the hourly pay budget tool to build employer cost; it cannot tell you what wage is lawful or competitive at your address.

The four-hour example: $120 before the first order

Assume the same $30 food order, menu and number of orders in both channels. There is no restaurant delivery charge, discount or refund in this first comparison. The hypothetical app charges 25%, with processing included and no other charges.

25% means 25 out of every 100 dollars. Type 25 ÷ 100 = 0.25 into a calculator. Then $30 × 0.25 = $7.50 per order.

For direct ordering with an employee driver, suppose:

• Four paid hours × $26 full employer cost per hour = $104. The $26 consists of $22 wage plus an illustrative $4 in employer costs; it is not a recommended wage or standard payroll burden.
• Delivery insurance, software and other fixed delivery costs allocated to this shift = $16. Replace this with quotes and a sensible allocation across your planned shifts. A new monthly bill is still owed if you cancel shifts.
• Total fixed delivery cost for the shift = $104 + $16 = $120.
• Direct payment processing at a fictional 3% plus $0.30 = ($30 × 0.03) + $0.30 = $1.20 per order.
• Vehicle/reimbursement cost, including return travel = $1.80 per order. This is a made-up total, not an IRS mileage rate or a reimbursement recommendation.
• Total direct variable cost = $1.20 + $1.80 = $3 per order.

Fixed means you pay that $120 for the staffed window even if no orders arrive. Variable means the cost changes as orders are added. At zero orders, the direct-delivery cost is still $120; cost per order is undefined because you cannot divide by zero.

Calculate the number of orders needed

Each order moved from the app to direct saves $7.50 − $3 = $4.50 before paying the driver's fixed shift cost.

Cost crossover = fixed shift cost ÷ savings per order.

$120 ÷ $4.50 = 26.67 orders. Since you cannot deliver two-thirds of an order, you need 27 whole orders for direct delivery to cost less under these assumptions. This is a delivery-channel comparison, not the break-even point for your entire restaurant.

Orders in 4 hours	App cost	Your driver + direct costs
10	$75	$150
20	$150	$180
27	$202.50	$201
30	$225	$210

The direct column is $120 + ($3 × orders). At 30 orders, the saving is just $15 for the shift. These rows show cost math only; they do not establish that one driver can handle that volume.

If your direct variable cost is equal to or greater than the app cost per order, there is no positive cost crossover with an added fixed shift cost. More orders will not fix that particular comparison. Recheck the service, prices and expenses.

Now check whether those orders can be delivered

Four hours contain 4 × 60 = 240 minutes. If one order takes 20 minutes from pickup through driving, parking, handoff and return, then 240 ÷ 20 = 12 deliveries. That is an optimistic ceiling before breaks, traffic surprises or time waiting for food.

To complete 27 orders, one driver would need to average 240 ÷ 27 = 8.89 minutes per completed order across the whole shift. Nearby orders delivered together can reduce time per order, but only if the routes, promised arrival times and food quality work. Don't assume every order can be paired.

Adding a second driver changes the calculation. In this example, another four hours at $26 adds $104. Fixed cost becomes $224 before any additional insurance or equipment. $224 ÷ $4.50 = 49.78, so the cost crossover moves to 50 orders. Two drivers doing single-order, 20-minute cycles can complete only 24 in four hours. Hiring another person does not rescue this example.

Measure actual routes during the hours you plan to offer delivery. Track the busiest 15-minute periods, not just the average night. Fifteen orders arriving together can overwhelm a service that looks fine across four hours. Set a delivery area and order limit you can honor; keep a priced backup for a driver absence or excess demand.

What if fewer people order directly?

The cost table assumes every order follows you. Customers who discover you through an app may never visit your website. Compare the money remaining from the orders each channel is likely to produce.

Suppose food and packaging cost $10 on each $30 order, and existing kitchen labor and overhead stay the same:

• App: $30 − $10 − $7.50 = $12.50 per order. Thirty orders leave $375.
• Direct with your driver: $30 − $10 − $3 = $17 per order before the fixed driver cost. Thirty orders leave (30 × $17) − $120 = $390.
• Lose just one direct order: (29 × $17) − $120 = $373, already below the app's $375.

This is contribution: money left to pay costs not included in the comparison. It is not take-home profit. Add any extra kitchen labor, refunds, marketing or other costs caused by the change. If the channels use different menu prices or food mixes, recalculate each side separately.

A delivery fee retained by the restaurant can change the result. Include its net revenue after applicable processing and other charges, then test whether customers still order at the full checkout price. Employee tips are not restaurant revenue. The U.S. Department of Labor's tipped-employee guidance explains tip protections and the need to follow more protective state rules. Have your payroll provider confirm the applicable wage, tip-credit and reimbursement requirements before staffing.

Copy this comparison before you choose

Use one sheet for each delivery window. Fill in:

1. Window and delivery area: ____
2. Expected app orders: ____; expected direct orders: ____
3. App contribution per order after food, packaging and all order costs: $____
4. Direct contribution per order before fixed driver costs: $____
5. Fixed driver and delivery costs for the window: $____
6. App result = app orders × app contribution = $____
7. Direct result = direct orders × direct contribution − fixed driver costs = $____
8. Deliveries physically achievable, including peak demand: ____
9. Extra setup cash needed and monthly commitments: $____

Get your insurer's written answer about the proposed drivers, vehicles and delivery use. Get actual ordering and courier quotes. Confirm who handles a missing order, a refund and a failed delivery. If the owner plans to drive, assign a cost to that time and check who runs the restaurant while the owner is away.

For a first restaurant, test a small delivery area during selected windows before promising delivery all day. Keep direct orders, app orders and courier-only orders separate in the POS. Compare similar shifts for money remaining, late orders, refunds and repeat customers. Keep the service that earns enough and delivers the food well; expand only when the numbers and the routes support it.

Use the free delivery profit calculator to check marketplace order contribution and the direct-pickup comparison. Use the worksheet above for the hired-driver decision; the calculator does not model driver routes or this shift-cost crossover. If the percentage steps are unfamiliar, start with Restaurant Math From Zero.

Sources and the question behind this guide

A recent public discussion about hiring delivery drivers instead of relying on apps asked what the numbers would show. A separate discussion about direct-delivery alternatives raised wait-time and service concerns. They displayed “10d ago” and “4d ago” when read September 12, 2026. Those questions informed this guide; forum opinions are not the source of its wage or pricing claims.

The SBA's break-even guidance explains fixed costs and per-unit contribution. The worked examples here apply that arithmetic to a delivery decision. Platform options and wage guidance were checked against the primary sources linked above on September 12, 2026. Your signed terms, current local requirements and measured operating results determine your answer.