On-demand delivery looks like a technology business and behaves like a logistics one. The app is the easy part. What decides whether the business works is order density in a small area, a driver network that turns up, and whether each delivery makes money after the driver, the fuel and the platform costs are paid.
Founders who scale before proving those three usually scale a loss.
Density beats coverage
A delivery business earns when a driver can complete several orders in a short radius. Spreading thinly across a city means drivers spend their time travelling rather than delivering, and the cost per order rises.
Start in one suburb and saturate it. Expanding into a second area before the first is dense is the most common and most expensive error in this model.
The driver network is the product
Customers experience the business as the person who arrives. Recruitment, vetting, training and retention of drivers therefore matter more than the interface.
Drivers leave when earnings per hour are unpredictable, which happens when order volume is thin. That makes density and driver retention the same problem rather than two.
Know the cost of one delivery
Work out what a single order costs you: driver payment, fuel, insurance, payment fees and your share of platform and support costs. Compare that with what you actually keep from the order.
If a delivery loses money, more deliveries lose more money. Growth does not fix negative unit economics, it accelerates them.
Merchants are the other half
Restaurants and retailers need a reason to join: incremental orders they would not otherwise get, or lower cost than delivering themselves. Commission has to leave them a margin or they leave.
Reliability is what retains them. A merchant whose orders arrive late absorbs the customer complaint, and they will stop rather than argue.
The obligations that come with it
Drivers engaged as employees bring full labour obligations, and the distinction between employee and independent contractor turns on how the relationship actually works rather than what the contract says. The requirements are set by the Department of Employment and Labour.
Customer data carries protection obligations, and vehicle and goods-in-transit insurance need to actually cover commercial delivery rather than private use.
Frequently asked questions
What decides whether a delivery business works?
Order density in a small area, a reliable driver network, and positive unit economics on a single delivery.
Why is density more important than coverage?
Because drivers earn and the business profits when several orders are completed in a short radius. Thin coverage raises cost per order.
What should I calculate before scaling?
The full cost of one delivery against what you keep from it. Scaling negative unit economics accelerates losses.
Why do drivers leave?
Unpredictable earnings per hour, which is caused by thin order volume. Driver retention and order density are the same problem.
What obligations apply to drivers?
Labour law where they are effectively employees, determined by how the relationship works in practice rather than by the contract wording.
Further reading
Originally published in January 2018. Updated September 2026 to explain what an on-demand delivery business actually requires rather than profiling one operator.
