Win restaurants first, then eaters. A delivery business has two customers, and eaters will not install an app with ten bland listings. Aim for 20 to 30 partner restaurants, including at least five names people already crave, in one compact area before you spend on any offer. Then promote through the places where people already order: housing societies, hostels, offices and WhatsApp groups.
The budget figures below are ballpark estimates for a one-town start. Platform rules, ad prices and commission norms change, so treat the numbers as planning tools and test small before you scale.
Your first 10 restaurants, before any customer
Restaurant owners decide to join in one conversation. They want orders, fast payouts and no surprises. Your pitch has to answer what they pay now, how they get paid and what they lose if you fail.
- Start with the top-selling kitchens in one dense area, not the longest list. Ask people on the street, at the nearest college or office park, and in local food groups where they order from.
- Visit between 3 and 5 pm when the kitchen is quiet. Do not turn up at 8:30 pm.
- Bring a one-page term sheet: commission rate, weekly payout day, what happens to cancelled orders, who pays when food is spilt, packaging responsibility and how to leave.
- Offer a reduced commission for the first two or three months, in exchange for a table sticker, a bill insert and a post to their own WhatsApp contacts.
- Sign the menu, not just the name: collect photos, prices, prep time per item and cut-off hours before you onboard.
- Name the weak slot: many restaurants have a slow afternoon or a late-night idle kitchen; a launch push in that slot is easier to promise and measure.
Ten committed restaurants are better than forty lukewarm ones. For comparison, see how restaurant online ordering works from the restaurant's side.
Where eaters are, and which channel suits which area
| Channel | How it works for delivery | Cost shape (estimate) | Best in |
|---|---|---|---|
| Housing societies and apartment WhatsApp groups | Resident welfare association (RWA) tie-up, stall at a society event, pinned menu in the group | A stall fee or sponsorship, plus a launch discount | Gated communities, townships |
| College and hostel marketing | Menu on hostel noticeboards, campus ambassadors, exam-night offers | Ambassador stipend or free orders; printing | Campus towns |
| Office parks and industrial estates | Lunch pre-order for floors or HR coordinators, group order links | Sample day and a sales visit | IT parks, factories, coaching hubs |
| Google Business Profile and local search | Appear for 'food delivery in [your town]' and area queries; see local SEO | Mostly time; photos and posting | Everywhere; slow but steady |
| Instagram and Facebook local ads | Geo-targeted to a 3 to 5 km radius with a first-order offer | Daily ad budget, test small | Cities with an active young base |
| Restaurant bill inserts and table stickers | QR code with a first-order code | Printing a few thousand cards | Everywhere your partners have footfall |
| Local influencers and food bloggers | Free delivery day or review order | Free meals and a small fee | Smaller towns where they are known |
Choose two channels in your first month, not seven. Your territory is small, and a stall at one large society can generate more first orders than a week of ads.
Launch pricing and offers that do not wreck the unit economics
A delivery order must pay for the rider, the payment fee, any refund and your overheads. A discount-led launch can bring a rush of orders and a rush of losses. Plan the offer on paper first.
| Offer | What it costs you | When it works | Watch out for |
|---|---|---|---|
| Free delivery on the first order | One rider cost per new customer | To get a first try in a new society | Abuse through multiple numbers; cap by phone and address |
| Flat discount on a minimum basket (say ₹100 off above ₹399) | Fixed rupees, but raises basket size | To lift average order value | Customers learn to wait for the code |
| Restaurant-funded offer | Cost borne by the partner, often for slow slots | Mid-afternoon or late-night menus | Restaurant needs to see sales to stay committed |
| Weekday combo with a hostel or office | Lower margin per meal but higher volume per rider | Dense lunch drops | Quality dips when kitchens rush |
| Referral reward on both sides | Paid only after the friend's first delivered order | After you have a happy base | Fraud with a second number; cap per customer |
- Know your delivery cost per order (rider pay, fuel, support share) before you choose a discount.
- Set a daily cap on discounted orders so a viral post cannot break the week.
- Avoid starting with a permanent price cut. A launch offer ends; a price that is always low trains customers to expect it.
- Keep the delivery fee visible. Eaters accept a clear small fee more readily than a hidden markup.
Turning a first order into a habit
The cheapest order you will ever get is the second one. A delivery business is mostly a repeat-order business, and repeat depends on what happens in the first 45 minutes of the first order.
- Hit the promised time on the first order. A late first order costs more than any discount can buy back. Quote a slightly longer window and beat it.
- Call the customer after the first order (or send a WhatsApp message) and ask one question: was the food hot and complete?
- Apologise with a specific credit, not a long message, when something goes wrong.
- Slot-based reminders: a Thursday evening WhatsApp for a weekend biryani, or an exam-week lunch message to hostels.
- Order-again shortcuts: a saved last order, a WhatsApp 'repeat my last order' reply or a one-tap button in the app.
- A simple loyalty cue, such as every sixth order's delivery free, rather than a points scheme that requires explanation.
- Measure the repeat rate at 30 days among customers in each society or hostel and put more effort where it is highest.
Privacy matters when you message customers. Ask for consent and offer an opt-out, and see our DPDP Act checklist for what the data protection law expects of you.
Using WhatsApp, your own channel and the open network
Many first orders in a small town are taken on WhatsApp, long before an app is installed. A catalogue and a clear order template is a valid first channel, and it gives you a customer list you own.
- WhatsApp Business catalogue with partner restaurants' top items and prices, and a standard order message with address, time and payment.
- Broadcast lists, not group blasts, to avoid being muted. Ask for opt-in at the first order.
- Your own site or ordering link so customers who find you on Google can order without a chat; the business website solution covers what that needs.
- ONDC, the Open Network for Digital Commerce, lets sellers and logistics providers reach buyers through other buyer apps. Check the network's current participation rules, fees and onboarding steps on its own site, because they change and we have not verified current terms.
- Aggregator listings for your partners: your restaurants may also be on big platforms; do not ask them to leave, but give them a reason to push their regulars to you, such as a better price on your channel.
Platform and network terms, including commissions, are not stated here because they change and we have not verified current figures. Read each platform's published terms before committing.
A 30-day plan and numbers to watch
| Week | Focus | Target to aim for (planning goal, not a benchmark) |
|---|---|---|
| 1 | Sign and photograph partners; train riders; test internally | 10 to 15 restaurants live, 5 to 8 riders |
| 2 | Soft launch to friends, staff and one society | 20 to 30 orders a day, to find bugs |
| 3 | Launch offer in one or two societies or a campus; Google profile live | First 100 customers; call 30 of them |
| 4 | Review delivery times, cancellations and complaints; fix the worst restaurant and the worst slot | A repeat rate you can measure by society |
Track four numbers daily: orders per rider per hour, average delivery time, cancellation rate and the percentage of first-time customers who reorder within 14 days. If the first is low, fix density; if the second is high, fix restaurant prep times; if the third is high, fix menu accuracy; and if the fourth is low, fix quality before spending more on offers.
The full playbook covers licences and the build plan, and what to buy and who to hire cover your kit and team. If you are weighing a single-kitchen model, see a cloud kitchen.
More on starting a food delivery business
- Start hereThe full playbook →Models, licences, startup costs and the steps to your first customer.
- Go deeperWhat to buy →Bags, bikes, phones, insurance and a dispatch desk, costed, and what to hold off buying.
- Go deeperWho to hire →Riders, dispatcher, support and restaurant sales: order, pay, vetting and contract types.
Questions people ask
How do I get restaurants to join a new delivery platform?
Visit top-selling kitchens in one dense area in the quiet afternoon, with a one-page sheet showing commission, weekly payout day and what happens to spilt or cancelled orders. Offer a lower commission for the first two or three months in return for a table sticker and a post to their regulars. Ten committed partners beat forty lukewarm ones.
Where do I find my first customers?
Housing societies, hostels and office parks are the densest sources. Pitch a residents' welfare association, sponsor a small stall, post your menu in resident WhatsApp groups and give a first-order offer. Add a Google Business Profile so people searching for delivery in your town find you. Start with two channels, not seven.
Should I offer big discounts to launch?
Only after you calculate your delivery cost per order. A capped first-order offer or a restaurant-funded slow-slot deal usually works. Heavy open-ended discounts bring a rush of orders and losses, and they teach customers to wait for codes. Set a daily cap on discounted orders so a spike cannot break your budget.
How do I get customers to reorder?
Deliver the first order on time and hot, then check in with a call or WhatsApp message. Make reordering one tap with a saved last order. Send timed reminders for patterns such as weekend biryani or exam-week lunches, and use a simple cue like free delivery on every sixth order. Track the 14-day repeat rate by area.
Is it worth joining ONDC as a small delivery operator?
It can widen your reach without building every side alone, but terms, fees and onboarding steps change, and we have not verified the current ones. Read the network's published participation rules and talk to a technology partner already connected before you commit. Treat it as an extra channel, not a replacement for local density.
