Skip to content
Restaurant Online Ordering

How to Set Up Online Ordering for a Cloud Kitchen

A setup guide for cloud kitchens adding an online food ordering system beside the marketplaces: brand pages, delivery-safe menu, riders and pricing.

Published · 7 min read

A cloud kitchen has no front door. Every rupee it earns arrives through a screen, and for most operators that screen belongs to a marketplace. This guide is for owners running one to five delivery brands out of a shared kitchen who want a second channel they control. By the end you will have a plan for the brand pages, the menu, the riders, the packaging and the pricing, and a realistic sense of how long each step takes.

Why cloud kitchens feel marketplace costs more than anyone

A restaurant with tables can absorb a marketplace commission because dine-in and walk-in takeaway carry no commission at all. A cloud kitchen cannot. If all of your revenue comes through aggregators, the commission is not a cost on one channel; it is a cost on the whole business.

The reported ranges are wide, and they depend on the contract. Business Standard reported in August 2024 that Swiggy typically charges restaurant commissions of 17 to 25 per cent. Verdict, citing NDTV Profit in September 2025, put Zomato’s range at 10 to 28 per cent. In the United States, Uber Eats publishes marketplace plans at 20, 25 and 30 per cent for delivery orders, and DoorDash publishes 15, 25 and 30 per cent (both checked on their pricing pages on 29 October 2026). Whatever your own rate is, it applies to the regular who orders your butter chicken every Friday just as much as to the first-time customer who found you by scrolling.

That is the case for a direct channel. Not to replace the marketplaces, which still do the discovery work, but to give repeat customers somewhere cheaper to order and to keep their details with you.

Step 1: Decide what the direct channel is for

Be specific. For a cloud kitchen, the direct channel usually has one job: capture the second, third and tenth order from people who already like the food. It is rarely a discovery channel, because nobody browses a kitchen they have never heard of.

Write the goal down in one line, for example: “Move 15 per cent of monthly orders from Brand A’s regulars to our own site within six months.” A target like that shapes every later decision, from which brand to start with to how aggressively you price.

Step 2: One ordering page per brand

Cloud kitchens often run several concepts from one line. Customers know the brands, not the kitchen, so each concept needs its own ordering page with its own name, colours and menu. A biryani brand and a burger brand should not share a homepage even if they share a fryer.

Start with the brand that has the strongest repeat behaviour. You can check this in your marketplace dashboard by looking at repeat-customer share per brand, if the platform exposes it, or simply by which brand gets the most “same again” orders in the kitchen’s memory. Launch that one first and learn before you roll the others out. The restaurant website builder page explains what a branded ordering page includes.

Step 3: Simplify the menu for delivery

The menu that works on an aggregator is often too long. Marketplace menus grow because adding items is free and might catch a search. On your own page you are talking to people who already know what they like, so trim.

Keep items that travel. A dish that is excellent at the pass and soggy after twenty minutes in a box will cost you a repeat order, which is exactly the order the direct channel exists to win. For each item, ask:

  • Does it survive 25 minutes in a sealed container?
  • Can the kitchen make it in the quoted prep time during a rush?
  • Does it need variants (size, spice level, base) or add-ons (extra raita, extra patty) to be ordered correctly?

Tuckly’s menu builder handles variants and add-ons with their own prices, and lets you set availability windows and switch items off mid-service when they sell out. Use the availability windows for brands that only run at certain hours, so a breakfast concept does not take orders at 9pm.

Step 4: Decide how the food gets there

Cloud kitchens live or die on delivery, and here the options are narrower than for a restaurant with walk-in trade.

Option Fits when Watch out for
Porter rider on demand (India, in cities Porter serves) Volume is uneven and you do not want riders on the payroll Porter’s public API currently covers two-wheelers only; the parts of the Tuckly workflow that automate booking and tracking are still being validated, so check the Porter integration page for what is live
Your own riders Steady volume in a tight radius, or late-night hours third parties avoid Idle time between orders; insurance and vehicle costs
Pickup A small share of customers who live or work nearby Rarely large for a cloud kitchen, but cheap to offer

Many kitchens end up with a mix: one rider on shift for the lunch and dinner peaks, Porter for overflow. Tuckly treats fulfilment as a per-order choice, so you can assign your rider to one order and request a Porter rider for the next. The delivery management page walks through the workflow and its limits.

Step 5: Set delivery areas and charges you can actually serve

Draw the delivery area around what a rider can reach in 15 minutes from the kitchen, not around where you would like customers to be. A direct order that arrives 50 minutes late because you accepted an address at the edge of the city does more damage than no order.

Decide what the customer pays per trip. If a Porter booking for a 4 km trip costs you a certain amount, you can pass it through, subsidise it on orders above a threshold, or build it into prices. The delivery fee calculator helps you test options against your average order value. Whatever you choose, show it before the customer reaches payment.

A cloud kitchen’s customers only ever meet the brand in a bag. That bag is your main promotional surface.

  • A printed card in every marketplace order with a QR code to the direct ordering page and a one-line reason to use it (“Order direct next time and skip the platform fee”). The QR code generator makes a print-ready code in a minute.
  • The ordering link in your Instagram bio and in every WhatsApp reply.
  • Your Google Business Profile, if the kitchen has one, with the direct link as the order button.

Check your marketplace partner agreement before adding inserts that mention other channels. Terms vary, and some restrict what you can include in orders fulfilled through the platform. The guide to promoting an ordering website covers each channel in more depth.

Step 7: Price the direct channel deliberately

You do not have to undercut the marketplaces, but customers need a reason to switch. Three common approaches:

  1. Same menu prices, no platform fee. The customer saves the marketplace’s platform fee and any surge; you save the commission. Simple to explain.
  2. Slightly lower direct prices. For example, a thali listed at ₹320 on the marketplace is ₹299 direct. You still net more than on a marketplace order after your payment gateway’s fee.
  3. Direct-only items or portions. A family pack that only exists on your page gives regulars a reason to look.

Remember the direct order still has costs: your payment gateway’s processing fee (Razorpay lists 2 per cent per domestic transaction plus GST on the fee, per its pricing page dated May 2026), the rider, packaging, and any subscription for the ordering software. The commission savings calculator puts your own numbers side by side.

Step 8: Staff the order screen like a station

Direct orders arrive on your own order screen, not in the marketplace tablet. Someone has to own it. In a shared kitchen the practical answer is usually the expediter or the person running the pass, with the screen mounted where they can see it and an audible alert they will not miss over the exhaust fans.

Set prep times you can keep. If the kitchen needs 18 minutes for a biryani during the dinner rush, say 20, not 12. Customers forgive a slightly longer quote; they do not forgive a missed one.

A realistic timeline

Week What gets done
1 Pick the launch brand, trim the menu, build it with variants and add-ons, decide delivery area and charges
2 Publish the ordering page, run ten test orders to friends, fix prep times and packaging
3 Print insert cards, update bio links and Google Business Profile, start asking for reviews with the direct link
4 onward Review orders weekly: repeat share, late deliveries, items that get removed from carts, and adjust

What to do next

Choose one brand, trim its menu to what travels, and set it up on a page you own. Keep the marketplaces running for discovery while the direct channel collects your regulars. The cloud kitchen solutions page describes how the pieces fit together for a multi-brand kitchen, including what is supported today and what is still being validated.

cloud kitchendark kitchendirect orderingdeliveryporter

Related reading

Get started

Create your ordering website

Tell us about your restaurant. We set up your online ordering system and help with your menu and delivery rules. No commission on direct orders.

What do you want to offer?

We reply by email. No spam.