Every delivery platform runs on the same three revenue lines, plus a few optional ones. The mistake at small scale is copying the mix a national aggregator uses, when your cost structure is nothing like theirs.
1. delivery fee
A percentage of the food subtotal, typically 15–30%. At maturity it is the largest line. At launch it is usually too low to cover delivery, and raising it too fast loses the outlets you recruited.
Math: 20% of a $10 basket is $2. If a delivery costs $2.50 all-in, delivery fee alone loses money on every order. That is not a pricing failure — it is a signal that the delivery fee has to carry the fleet.
2. Delivery fee
Charged to the customer per order, often tiered by distance, basket size or time of day. This is the line that should pay for riders. Most new platforms underprice it because they are afraid of losing the order, then discover the loss per delivery compounds with volume.
Rule of thumb: the delivery fee plus delivery fee should cover rider cost, payment fees and support, with something left over. If it does not, no amount of growth fixes it.
3. Surge and small-order fees
Peak-hour multipliers and minimum-basket penalties. Politically unpopular, operationally necessary. A $1 fee on orders under $8 removes the orders that lose money.
4. Advertising and placement
Restaurants pay to appear at the top of a category or in a banner. This only works with demand — sell it once an outlet can see incremental orders, not before. At small scale, bundled "featured" placements are easier to sell than CPC.
5. Subscriptions
A customer pays a monthly fee for free delivery. It stabilises frequency and gives you predictable revenue, but it only works when the average subscriber orders more than the fee's cost in delivery. Model it before you launch it.
6. outlet SaaS
Some platforms charge restaurants for the tablet, menu photography or their own branded ordering page. This is a clean line because it is a service, not a tax on orders — and it is what an agency-style operator can sell alongside the platform.
What the P&L actually looks like
| Per $10 order | Amount |
|---|---|
| Customer pays (food + delivery) | $12.00 |
| outlet keeps (80% of food) | -$8.00 |
| Rider cost | -$2.50 |
| Payment gateway (~2%) | -$0.20 |
| Support and refunds provision | -$0.30 |
| Contribution | $1.00 |
That $1 is your entire margin — before marketing, before the app store, before your own time. It is why delivery businesses obsess over basket size and batching: both increase contribution without increasing the fixed cost of a trip.
The uncomfortable truth
Software does not make a delivery business profitable. Density does. Every extra order in the same radius lowers the cost per delivery; every new radius raises it. Price like that is true, and the numbers get easier.
A contribution model you can defend
Revenue lines are not the same as profit. Build a contribution model per order: commission + delivery fee + small-order fee + advertising revenue − rider cost − payment fee − packaging/support − refunds. Then subtract fixed costs (hosting, salaries, support, marketing) to get monthly profit.
| Line | Example |
|---|---|
| Average order value | $25 |
| Commission (15%) | $3.75 |
| Delivery fee | $2.00 |
| Rider cost | −$2.50 |
| Payment fee (2.5%) | −$0.63 |
| Packaging/support | −$0.50 |
| Contribution | $2.12 |
At 1,000 orders a month, that is $2,120 of contribution before fixed costs. If fixed costs are $1,500, the platform makes $620. A 5% refund rate can erase most of it. This is why order count alone is a misleading metric.
The delivery-fee trap
Many platforms subsidise delivery to win customers, then discover the subsidy is the largest line in the P&L. A “free delivery” offer is a marketing expense, not a revenue model. Set a delivery fee that covers the rider cost in the median order, and use free delivery only as a capped promotion or a loyalty benefit. If the fee cannot cover the rider, the commission must; if neither can, the order is losing money.
Small orders are the hardest. A $10 order with a $2.50 rider cost and a 15% commission leaves $1.50 before payment fees. A minimum basket or small-order fee is not anti-customer; it is what makes the economics work.
What to fix before adding a revenue stream
- Make the core order flow reliable; a new revenue line on a broken checkout is wasted.
- Know contribution per order by zone and vendor.
- Reduce refunds and failed deliveries before adding ads or subscriptions.
- Get restaurant settlements right; vendors who distrust payouts will not promote you.
- Only then add promoted placement, subscriptions or a loyalty programme.
Three realistic scenarios
Thin market: 300 orders/month, $18 basket, 15% commission. Contribution may be negative after rider and payment costs. The fix is density or a higher basket, not more software.
Working city: 1,000 orders/month, $25 basket, 15% commission, $2 delivery fee. Contribution is positive if rider cost stays near $2.50 and refunds stay below 3%.
Dense market: 3,000 orders/month, batching and higher basket. Contribution per order may fall slightly but total contribution rises. This is where a fixed software licence and efficient dispatch pay off.