Small delivery operations fail for the same three reasons: they buy software they cannot operate, they price delivery below cost, and they launch across too wide an area. None of those are technology problems, which is why buying more technology does not fix them.
Here is a launch plan sized for one owner, one kitchen and a handful of riders.
1. Start with one delivery radius
Two to three kilometres. Draw it on a map and count the customers inside it. A small operation that delivers hot food in 25 minutes inside 2 km will beat a bigger menu that arrives cold from 8 km away. Expand the radius only when the inner ring is consistently busy.
2. Choose the cheapest channel that works
| Option | Up front | Ongoing | Best for |
|---|---|---|---|
| Own website + WhatsApp | low | hosting | testing demand |
| Productised platform | one-time licence | ~$12/mo server | launching properly |
| Hosted SaaS | low | $99–$499/mo | no technical capacity |
| Custom build | $25,000+ | maintenance | not a small business |
For most small operators the practical answer is a productised platform: $299 once, installed on a $12 VPS, with the source code so you are not locked in.
3. Price delivery honestly
Work out your true cost per drop: rider time, fuel, bag wear, and the failed-delivery rate. If that number is $2.50 and you charge $1, every order loses $1.50 — and volume multiplies the loss. Charge the real number, or set a minimum basket that makes the trip worthwhile.
A minimum order of $8–$10 plus a $1.50–$2.50 delivery fee is a normal small-market structure. It will cost you some tiny orders, and those are the ones that were losing money.
4. Skip the feature list
Do not buy for loyalty points, AI recommendations or gamification. Buy for: accurate menu, order alerts, zone pricing, payments, pickup and reorder. Everything else can wait until you have a problem it solves.
5. Get the first 100 orders without ads
- QR code on every table and bag — cost: printing. Conversion: high, because the person is already eating your food.
- Google Business Profile → "Order online" pointing at your storefront, not an aggregator.
- A pinned Facebook post with the link and the delivery area.
- A first-order code valid only on your own channel.
- Ask every regular at the counter for the order — the single highest-converting channel there is.
6. Operate one shift before you market
Place ten orders yourself: one cancellation, one refund, one wrong address, one item out of stock, one order during peak. Fix what breaks. Then open the doors. Launching to 200 customers with a broken refund flow costs more than a week's delay.
7. Measure three numbers only
Orders per week, repeat rate, and delivery time from "ready" to "delivered". If orders grow and repeats do not, the food or the timing is the problem, not the software. If repeats grow and orders do not, you need reach, not features.
What "small" gets you
Speed and consistency. A small operation can answer the phone, know the regulars and deliver hot food. Software should protect those advantages, not replace them with a support queue.
A realistic small-business budget
A small delivery business does not need a five-figure platform. It needs a reliable order flow, a delivery radius it can serve, and a cost structure that survives a quiet week. A realistic first-year budget has four lines: software, hosting/integrations, rider cost and marketing.
| Line | Lean launch | What it covers |
|---|---|---|
| Software | $299 one-time | Apps, admin, website, source code |
| Hosting + services | $30–$100/month | VPS, SMS, maps, email, backups |
| Rider cost | $2–$4 per delivery | Per-delivery pay or shifts |
| Marketing | 10–20% of early revenue | Local ads, referrals, launch offers |
If you cannot cover rider cost and marketing from contribution, the business is not ready to scale. A cheaper licence will not fix a negative contribution per order.
Single vs multi-vendor for a small operator
Start single-vendor if you own the restaurant or have one anchor brand. You keep the full basket margin and the operation is simpler. Add vendors only when you have repeat demand and can support onboarding, settlements and menu quality. A small multi-vendor marketplace with 10 restaurants spread across a city is harder than a single brand with a tight delivery radius.
The first 30 days
- Choose one 3–4 km radius and 25–40 menu items.
- Set a delivery fee that covers the median rider cost.
- Run 20 test orders with staff and friends; fix packaging and prep times.
- Launch with one payment method and cash on delivery if the market needs it.
- Ask every customer for a review and a referral.
- Review contribution per order weekly, not revenue.
When to upgrade
Add a second zone when the first zone is consistently on time and contribution-positive. Add vendors when you have more demand than your own menu can serve. Add loyalty when you have repeat customers, not before. Add a rider fleet when per-delivery pay cannot cover peak demand. Each upgrade should solve a measured problem, not a fear of looking small.
Three numbers to watch
Contribution per order after rider, payment and packaging. Repeat order rate within 30 days. On-time delivery rate. If those three are healthy, the business can grow. If they are not, more orders will only amplify the problem.