A restaurant shopping for delivery software is usually not shopping for software. It is trying to stop handing a third of every order to an aggregator, and it has been quoted five figures by an agency. That framing matters, because it changes what you should buy.
You do not need a multi-outlet platform. You need a menu, a cart, a kitchen queue, a delivery rule and a customer list you own. Anything beyond that is optional for the first year.
The commission maths, done properly
Work out what the aggregator actually costs you per month, not per order. A restaurant doing 400 orders at a $12 average basket and paying 28% hands over roughly $1,344 a month — about $16,000 a year. That is the budget you are reallocating, not the software price.
Now the honest part: the aggregator also brings demand. Moving to your own channel does not move that demand with you. The first three months are marketing work — a Google Business Profile that points at your ordering page, a QR code on every table, a card in every bag, a pinned post on your Facebook page. If you are not prepared to do that, the commission is buying you something real.
The usual healthy outcome is both channels: aggregators for discovery, your own app for regulars. Repeat customers are where the margin is, and repeat customers are the ones who will install your app.
What the software has to do
- Keep the menu accurate. Sold-out items must disappear the moment the kitchen runs out. A customer who orders something you cannot cook does not come back.
- Never lose an order. Sound alerts, a visible queue, and an escalation if nobody accepts in ninety seconds.
- Price delivery by area. A flat fee either overcharges nearby customers or loses money on the far ones.
- Take the payments your customers use. Card, wallet and cash on delivery, with the merchant account in your name.
- Let people reorder in two taps. The highest-leverage retention feature in food ordering, and the cheapest to build.
- Show order status. "Preparing" and "on the way" remove most of the phone calls.
What you can ignore for now
Loyalty wheels, gamified referrals, AI recommendations, in-app stories, multi-language menus. None of it matters before you have repeat customers, and all of it adds surface area to break. Add features when a specific problem appears, not because a competitor's app has them.
Pickup is a feature, not a compromise
Most restaurants underuse pickup. It has no rider cost, no delivery radius and no timing risk, and it converts customers who live close but do not want to pay a delivery fee. Turn it on from day one and price it as free or near-free.
Delivery: three models
Your own staff delivering is the cheapest way to start and the least scalable. Platform riders cost a shift whether or not orders arrive. Third-party fleets price per drop. Start with your own staff inside a 2 km radius, then add capacity when the radius is consistently saturated — not before.
What it should cost
A productised platform is a one-time licence plus hosting — Foodomaa is $299 and about $12 a month for a VPS. A custom build for three apps starts around $25,000. A hosted SaaS is $99–$499 a month forever. For one restaurant, the custom build is almost never the right first move; the software is not your differentiator, the food is.
Questions to ask before you buy
- If I stop paying, does my storefront stay online?
- Can I reconcile a payout against my own sales report?
- Who owns the Play Store listing — me or the vendor?
- What happens when a rider goes offline mid-delivery?
- Can I change the delivery fee without a support ticket?
If the answers involve a licence server, a tenant slot, or "our team will handle it", you are renting. Renting is fine for a test and expensive for a business.
Aggregator vs own channel
Aggregators bring demand but charge commission and own the customer relationship. An own ordering channel gives you the customer data, the menu margin and the ability to run your own offers, but you have to build demand. Most restaurants need both: aggregators for discovery, the own channel for repeat customers.
| Aggregator | Own channel | |
|---|---|---|
| Demand | Borrowed | Built |
| Commission | 15–30% | No platform commission |
| Customer data | Limited | Yours |
| Delivery | Platform riders | Your riders or self-delivery |
| Marketing | Platform + yours | Yours |
Feature checklist
- Menu with modifiers, availability and scheduled hours.
- Delivery zones, minimum basket and fees.
- Pickup and self-delivery options.
- Payment methods used in your market.
- Order acceptance, prep time and rider handover.
- Customer database, offers and repeat-order tools.
- Reports by item, channel and daypart.
Cost and commission
A $299 licence plus hosting can be cheaper than a year of aggregator commission on repeat orders, but the licence does not bring demand. Budget for local marketing, a photographer, packaging and a rider or delivery partner. If you cannot fund demand, the own channel will sit empty while the aggregator keeps the customers.
A 30-day launch plan
- Pick 20–30 menu items that travel well and cost them with packaging.
- Install the software, set zones, fees and payment methods.
- Photograph the top 10 items and write honest descriptions.
- Run 20 test orders and fix prep-time and packaging issues.
- Launch to existing customers with a QR code, WhatsApp list and a first-order offer.
- Review contribution and repeat rate weekly.
When to keep the aggregator
Keep the aggregator for discovery and new customers, and use the own channel for regulars, larger baskets and profitable menu items. Do not try to beat the aggregator on every order; use it as a acquisition channel and move repeat demand to your own app over time.
The aggregator math
If an aggregator charges 25% on a $20 order, the restaurant receives $15 before food cost. On the own channel, the same order may net $19 after payment and delivery costs. The difference is $4 per order. At 500 orders a month, that is $2,000 — enough to pay for a licence, hosting and a rider shift. The catch is that the aggregator brings demand; the own channel must earn it.
Menu and packaging for delivery
Choose items that travel well, cost them with packaging and price them for commission. A dine-in menu is not a delivery menu. Remove leaky, delicate or temperature-sensitive items unless you can package them reliably. Use modifiers to protect margin and reduce support calls: “extra sauce”, “no onions”, “make it spicy”.
A 30-day launch plan
- Week 1: choose 20–30 delivery-safe items and cost every one.
- Week 2: install the software, set zones, fees and payments.
- Week 3: photograph the top 10 items and load the menu.
- Week 4: run 20 test orders, fix packaging and launch to existing customers.
When to add your own riders
Add riders when delivery quality is the constraint and order density can keep them busy. A third-party fleet is easier at launch; an owned fleet gives more control at scale. Start with a hybrid: own a small core fleet, use partners for overflow, and measure cost per delivery and on-time rate every week.