US Indian Restaurant Direct Ordering Playbook

This playbook is for the owner of an Indian restaurant, tiffin service, sweet shop or catering kitchen in the United States whose customers are mostly South Asian families, and who is tired of watching a quarter of every order go to a delivery app. By the end you will have the fee math for your own order volume, a direct ordering channel on WhatsApp and a store page that your community will actually use, Stripe connected and billed in dollars, a tiffin and catering subscription model with deposits and cut-offs, a local marketing plan built around temples, offices and community groups, and a review routine that keeps your Google listing ahead of the apps. The belief it will change is that direct ordering is a thing you add after the apps have made you busy. For an Indian restaurant in America the order is reversed: your customers already find you through community and already live on WhatsApp, so the direct channel is the main road and the apps are the signboard on the highway. Searching for Indian restaurant online ordering USA options returns a page of app listings; this is the guide to the channel you own.

Why Indian restaurants in the US are built for direct ordering

Most American restaurants struggle to move customers off a delivery app because those customers found them on the app and have no other relationship with them. Indian restaurants are different in four ways, and each one is an advantage you are probably not using.

• Discovery is community, not search. Families in Edison, Fremont, Sugar Land, Naperville or Irving hear about a restaurant from a temple, a cricket group, a colleague or a parents' WhatsApp group. Word of mouth works without an app.
• Orders repeat. A family that likes your Sunday thali orders it most Sundays. Repeat orders are exactly where a 25 percent commission hurts most and is least needed.
• Tickets are large. Party trays, Diwali sweets, Navratri thalis and office catering run from $150 to several hundred dollars, and the apps handle them badly anyway.
• WhatsApp is already installed. Many South Asian households in the US use it daily to talk to family abroad, so an ordering number needs no explaining and no download.

Put those together and the strategy writes itself: keep the apps for the customer who has never heard of you, and move every repeat customer, every large order and every subscription to a channel where you pay a flat fee instead of a percentage. The commission-free ordering guide for Indian restaurants in the USA is the short version of this playbook; the chapters below go deeper with numbers and checklists.

Delivery app fee math in the US

Delivery app commission in the US is a percentage of each order's food subtotal, and the percentage depends on which partnership tier you choose, whether the order is delivered by the app's driver or picked up, and whether it came through the marketplace or an ordering page the app built for your website. The headline numbers are easy to find. What costs you money is everything around them.

How the tiers work

Check each app's current partner pricing page before signing; plans and rates change. The structural point matters more than the exact percentage: on the entry tier you show to fewer customers and sit outside the subscription programme, so most owners feel pushed to the middle tier to stay visible. The plan that keeps you competitive inside the app is the one that takes a quarter of every order.

What a $30 order nets you on each channel

This is an illustrative example built from typical numbers, not a named merchant.

• Marketplace, middle tier at 25 percent: commission $7.50. Food cost at about 30 percent of menu price is $9, packaging about $1. Left to cover labour, rent and everything else: about $12.50.
• Own WhatsApp and store page channel: flat platform fee $0.50 plus card processing at the standard rate of 2.9 percent plus 30 cents, about $1.17. Same food and packaging. Left: about $18.33.
• Difference per order: about $5.80. On 400 repeat orders a month, roughly $2,300. Over a year, the cost of a second cook.

The charges outside the commission line

• Promotions you fund. First-order discounts and "spend $25, get $5 off" offers are typically paid by the restaurant. A $5 promotion on a $30 order is another 17 percent gone.
• Sponsored placement. Paying to appear higher in the app's search is advertising spend on top of commission.
• Error charges. Missing or wrong items lead to customer refunds, and who bears the cost varies by situation. Track your error rate; it is a line in your fee bill.
• Tablet and hardware fees, small but recurring, depending on your agreement.
• Menu price parity rules, which have varied by plan and region over time. Read your agreement before listing higher prices on the app to offset commission.
• Sales tax handling. In most states the apps act as marketplace facilitators and remit sales tax on their orders. Not a fee, but it changes how you reconcile payouts, and on your own channel the tax is your job.

A month of orders, side by side

Illustrative: 500 orders a month at a $30 average, $15,000 in food sales.

The per-order breakdown by app, and what the pickup, storefront and flat-fee delivery options really save, are in US food delivery app commission fees for restaurants and delivery app fees in the US per order. The decision framework for a small restaurant weighing the two is in delivery apps versus direct ordering for small US restaurants.

Lowering the app bill without leaving

• Audit your tier every quarter. If the middle tier is not measurably producing more orders than the entry tier would, downgrade. Ask your account representative for the data.
• Push pickup inside the app. Every order that moves from delivery to pickup saves most of the commission.
• Switch off open-ended promotions. Run one for a new dish or a slow Tuesday, then stop.
• Put your own ordering link in front of every customer the app sends you: a card in the bag, a sign at the counter, the link on your Google listing. You may not message the app's customers using the app's data, which the platforms' terms generally prohibit, but you may make your own channel visible everywhere.
• Report net revenue per channel monthly. An app order and a direct order are not worth the same.

Delivery once the orders are yours

Start with pickup, which many Indian restaurant orders already are, especially the large family ones; set collection slots so the kitchen is not hit with ten party trays at 6pm. Add a part-time driver for a three-mile radius during dinner once volume passes a few dozen orders a night; it usually costs less per order than any third party. For the long-distance orders you do not want to drive, the apps and on-demand delivery partners will deliver orders you took yourself for a flat fee per delivery rather than a commission. Set delivery zones with charges by distance so the customer twelve miles away pays a fair fee rather than being refused or subsidised.

WhatsApp adoption among diaspora customers

The question American restaurant owners ask is whether customers will order on WhatsApp. For an Indian restaurant the better question is why they are not already doing it, because many of your customers message you there now, and the problem is that chat does not scale. By Friday evening someone is copying addresses into a notebook and asking for a screenshot of a bank transfer.

Why the number spreads on its own

A WhatsApp business number is saved like any contact and forwarded like any message. When a family in an apartment complex near a tech campus finds a tiffin they like, the number goes into the building's WhatsApp group that evening. A temple volunteer who ordered prasad trays shares it with the committee. The number travels through the same groups that already carry recommendations; an app link does not, because the recipient has to install something. The WhatsApp Business in the US guide covers who uses it in America and why the share is growing.

What the ordering flow should look like

1. The customer messages your business number, or taps a link from your Google listing, Instagram bio or a card in the bag.
2. A bot replies within seconds with a greeting and the menu, in English or Hindi, and the customer picks items from a list instead of typing them.
3. They choose pickup with a collection time, or delivery with an address inside your zones, or a scheduled slot for a later day.
4. They pay by card on a Stripe checkout, or choose cash on pickup if you allow it for regulars.
5. The order lands in your dashboard and the staff app with everything filled in, and the customer gets status updates in the same chat: accepted, preparing, ready for pickup or out for delivery.

Nobody on your side types an address, asks for a screenshot or calls back to confirm. That is the whole difference between taking orders on WhatsApp and running an ordering channel on WhatsApp.

Match the channel to the customer

Both WhatsApp and the store page feed one dashboard, so the kitchen handles a group-chat order and a browser order the same way. Do not build a custom app. Adoption is poor, the build is expensive, and your customers already have WhatsApp and a browser.

Messaging consent and cadence in the US

US rules on business messaging are strict about opt-in. Collect consent at the first order, send campaigns only to customers who opted in, give an easy way to stop, and honour every stop the same day. WhatsApp's own business policies apply on top and are the same worldwide. A realistic cadence for an Indian restaurant is one message a week, usually Thursday with the weekend special or thali menu, plus festival pre-orders and the occasional match-day snack box. The customers who block restaurants do it because of daily promotions, and a blocked customer is worse than a customer on the delivery app.

Stripe setup and USD plans

In India a direct ordering channel collects UPI and card payments through Razorpay. In the US the equivalent is Stripe: a payment processor that collects card and digital wallet payments securely, takes its fee, and pays the rest to your business bank account on a schedule. You do not write code to use it. You create an account, verify the business, connect it from your ordering platform's settings, and the checkout step on WhatsApp and the store page runs on it.

Stripe fees on a food order

Stripe's standard US online pricing, as listed on its website at the time of writing, is 2.9 percent plus 30 cents per successful domestic card or wallet charge, with no monthly or setup fee. On a $32 order that is about $1.23; on a $180 catering tray, about $5.52. International cards, currency conversion and instant payouts cost extra, disputes carry a fixed fee that is generally returned if you win, and a refund does not return the original processing fee. Very small orders carry a proportionally higher cost because of the fixed 30 cents; a $6 chai-and-samosa order pays close to 8 percent all in, which is one reason to set a sensible minimum for delivery.

Setup checklist

• Sign up with a business email and choose the business type: sole proprietor with a Social Security number, or LLC or corporation with its EIN. The legal name must match the IRS registration exactly.
• Enter the address, phone, ordering page URL and a plain description such as "Indian restaurant, online food orders for pickup and delivery". Stripe assigns a merchant category from it.
• Add the business checking account for payouts.
• Verify the owner's identity; have a photo ID ready in case Stripe asks.
• Set the statement descriptor to your restaurant's name, so customers recognise the charge and do not dispute it.
• Enable digital wallet options; they raise completion on mobile orders because nobody types a card number.
• Connect to your ordering platform through Stripe's own login and approval flow.
• Place a $1 test order from your own phone, confirm it appears in the dashboard with your name on the charge, then refund it.
• Confirm the payout schedule and the bank account before the first weekend.

The whole process takes under an hour with the documents ready. Payouts for established accounts typically arrive about two business days after the charge; new accounts often wait longer for the first one while Stripe verifies the business. Set payouts to daily so each deposit matches roughly one day of orders, match Stripe's payout report to the dashboard's revenue for that day, and book Stripe fees as a cost line rather than netting them against sales. The full walk-through, including disputes and failed payments, is in Stripe for restaurants.

Sales tax is yours on a direct channel

Prepared food is taxed differently from state to state, and in some states by city. On marketplace orders the app usually collects and remits as a facilitator; on your own channel you charge the right rate and remit it. Confirm the rate and which items it applies to with your accountant before you set menu prices, and make sure the ordering platform applies it per order. Stripe moves the total; it does not decide the tax.

USD plans and what a month costs

On Tezzu, orders are unlimited on every plan and the per-order fee is flat: a $12 lunch bowl and a $180 catering tray both cost $0.50. Plan charges and per-order fees are billed to your card in dollars through Stripe, and customer payments are collected through your Stripe account with each charge confirmed back to the order automatically. Illustrative: a restaurant doing 400 direct orders a month at a $30 average, $12,000 in sales, pays $69 plus $200 to Tezzu, $269, and about $468 to Stripe in processing, so about $737 all in, a little over 6 percent of sales. The same $12,000 on a marketplace at 15 to 30 percent is $1,800 to $3,600 before promotions. Setup takes about 30 minutes: connect your existing +1 number through Meta's embedded signup, build the menu in dollars, set pickup points or zones in miles, connect Stripe, and send yourself a test order. The details are in Tezzu for US merchants. If you are comparing this with POS-first systems that bundle their own processing, POS alternatives for small US restaurants sets them side by side.

Tiffin and catering subscriptions

Two product lines carry most of the direct revenue for an Indian kitchen in the US: the weekday tiffin for students, working couples and visiting parents, and the catering tray for pujas, graduations, cricket finals and office lunches. Both are large, repeat and badly served by the apps, which is exactly what a direct channel is for.

Tiffin: sell plans, not meals

Customers compare a tiffin with two things: a $15 to $20 restaurant delivery after fees and tips, and the $5 to $7 it costs them to cook. Price in the middle. A standard vegetarian lunch of dal, one sabzi, rice, three rotis and salad commonly sits between $9 and $14 depending on the city, with non-vegetarian meals $2 to $4 more. Sell weekday plans at a discount to single meals and collect before the month starts.

• A 5-day trial week at the single-meal price, so a nervous customer can test the food without a $200 commitment.
• A 20-lunch monthly plan at, say, $220 instead of $260, paid up front by card.
• A four-week menu rotation so nobody sees the same Tuesday twice in a month, scheduled by weekday so the right menu shows on the right day without re-typing it.
• A skip rule in writing: cut-off the evening before, a meal credit carried forward rather than a refund.
• Pickup points over door delivery. The leasing office of a large complex, an Indian grocery store that holds a crate for an hour, a temple on weekends. Thirty meals in one stop instead of thirty stops, with door delivery at a $2 to $3 premium for those who want it.

Illustrative: 60 subscribers on the $220 plan is $13,200 a month. Food at $4.50 a meal for 1,200 meals is $5,400, a commissary kitchen slot and insurance about $1,000, a part-time driver and fuel about $1,200, packaging about $500. That leaves roughly $5,000 before your own time, and the same kitchen slot can usually take the count to 100 before anything breaks. If you sell the monthly plan as one bundle paid up front, it is one order on a per-order platform; if customers order day by day, each day is an order. Know which model you run before you price the plan. The licensing, which is where most desi tiffin businesses go wrong because cottage food laws do not cover hot meals, and the growth steps from 25 to 200 meals, are in Indian tiffin service in the USA. The fitness-customer variant, with macro-counted boxes and weekly pickup, is in meal prep business in the US.

Catering: quotes, deposits and pickup windows

Anyone who orders a party tray should do it on your channel, because the apps are bad at large orders and because once a family has ordered catering direct, their weekly order follows. The flow:

1. A catering menu with per-head prices on the store page: a vegetarian spread at $12 to $16 a head, non-vegetarian at $16 to $22, with minimum counts and 48-hour notice stated plainly.
2. A quote on WhatsApp within the hour, with the head count, the dishes, the total, the deposit and the pickup or delivery window.
3. A 50 percent deposit by payment link to confirm, the balance before pickup. The order is confirmed when the deposit lands, not when the customer says "sounds good".
4. A pickup window on the day, with the trays labelled and the chafing dishes counted, or delivery for a stated fee inside your zones.
5. A thank-you and a review request the next day, and a note in the customer's record for the same event next year.

Illustrative: a Satyanarayan puja for 40 guests at $14 a head is $560. The deposit link for $280 goes out with the quote; Stripe processing on the full amount is about $16.54; the platform fee is $0.50. On a marketplace, the same order would carry $140 in commission if the app could handle it at all. The quote wording, what to do when a buyer goes quiet, and the GST-equivalent paperwork for corporate buyers are in catering orders on WhatsApp.

The diaspora festival calendar

Festivals in the US run on the Indian calendar and the American weekend. Diwali sweets and gift boxes go on pre-order three weeks ahead with pickup on the two days before, because many families celebrate on the nearest Saturday. Navratri gets a nine-day fasting menu. Pongal, Onam, Eid, Holi and Ganesh Chaturthi follow your community's mix. Match-day snack boxes during India cricket fixtures, with a pickup slot before the toss, sell well in any city with a large diaspora. Each one is a scheduled order with a cut-off and a prepaid checkout, which is the only way to cook the right quantity.

Local marketing to temples, offices and community groups

The Indian community in any American city is organised, online and offline, and that organisation is your marketing channel. A table with sweets and a QR code at a Diwali mela does more for direct orders than a month of app ads, and it costs a tray of laddoos.

The channels and what to do in each

The card in every bag

The single highest-return tactic is also the cheapest. A printed card in every bag, including the bags that go out through the delivery apps: "Order direct on WhatsApp, skip the fees, 10 percent off your first direct order", with a QR code to your number and one to the store page. A first-order coupon that applies automatically from the link means nobody types a code. The wording and timing for telling customers to skip the apps without breaking any platform's rules are in order direct campaigns.

A monthly rhythm

• Weekly: the Thursday menu message to opted-in customers, one message, one link.
• Monthly: one community event with a table, a tray and two QR codes; one office pitched for a lunch programme.
• Each festival: pre-orders opened three weeks out, a reminder two days before the cut-off, a greeting on the day.
• Every 45 days: a short win-back message to anyone who has gone quiet, with free pickup or a flat $5 off, then stop after two tries.

On Tezzu, campaigns go out from the same number customers order on, paid from a prepaid wallet, with the store link and the pre-order items already set up; Click-to-WhatsApp ads on Instagram and Facebook that open a chat with your number are available on the Growth plan.

Reviews and your Google profile

When someone searches "Indian food near me" in your suburb, Google shows a map with three or four restaurants, their ratings and their review counts. That map decides who gets the walk-in, the call and the direct order. Your rating inside a delivery app only matters to people already inside that app, where you pay commission on whatever they order. Reviews on Google are an asset you own.

Set the listing up to take orders

• Claim and verify the profile if you have not; it takes a postcard or a phone verification.
• Put your own ordering page as the order link. Google often fills the food ordering section with app links by default; add yours and put it first where the interface allows.
• Set the menu link and the website field to the same ordering page, so every button leads somewhere that takes orders.
• Where the chat option is available for your profile, set your WhatsApp business number as the chat destination; otherwise put a wa.me link in a post and in the description.
• Write the order instruction into the business description: "Order direct on WhatsApp at [number] or at [store link]. Pickup in [town], delivery to [areas]."
• Post every week or two with a dish photo and the order link; answer the questions in the Q&A section and seed one yourself: "Can I order on WhatsApp?"

A chat from Google is a hot enquiry with a short fuse; the person is comparing three listings and orders from whoever answers first, so the bot's instant greeting with the menu link is what converts it. The step-by-step, including how to count orders from the listing, is in the Google Business Profile and WhatsApp order button guide.

Ask for reviews at the right moment

Restaurants with hundreds of reviews are rarely better than the ones with twenty. They ask more consistently and make it one tap. Get the short review link from your profile dashboard, test it on a phone signed into a different account, turn it into a QR code for the bag and the counter, and keep the same link everywhere. Then ask at the moments that get a yes: thirty to sixty minutes after a delivery is marked delivered, when a customer says "food was great" in the chat, after a good rating, from a person at the end of a dine-in meal, and after a resolved complaint.

Two rules keep the listing safe. Google prohibits any incentive for a review, including discounts and free items, and it treats selective solicitation, asking only the happy customers, as a violation. Ask everyone for feedback, reply to all of it, and make the Google request part of a normal thank-you flow. Reply to every review within a day; the reply is visible to the next customer and does more for your rating than any amount of screening. The channels compared and a weekly routine are in how to get more Google reviews for your restaurant.

Owned reviews after every order

A rating collected on your own channel after delivery does two jobs. It tells you about a packing or quality problem before it becomes a one-star review, and it tells you who to send the Google link to, since you have their number because the order came through you. On Tezzu, order reviews and ratings are requested on WhatsApp after delivery and shown in the dashboard, and the customer's history sits next to them, so the daily review-ask is not a job someone forgets when the kitchen is busy.

Your 90-day plan

Frequently asked questions

Will American customers order from an Indian restaurant on WhatsApp?

Your South Asian customers almost certainly will, because many use WhatsApp daily and the number spreads through the community groups they already belong to. For US-born customers and non-Indian colleagues, a store page in the browser covers them with the same menu and card checkout, and both channels land in one dashboard. Most Indian restaurants find WhatsApp adoption high without any persuading.

What is the cheapest way for an Indian restaurant in the US to take online orders?

A direct channel with a flat per-order fee, such as WhatsApp ordering with a store page, is usually the lowest cost per order once you have regular customers: a monthly subscription, a small per-order fee and standard card processing, with no percentage commission. Delivery apps remain useful for customers who have never heard of you, so most restaurants keep one tier for discovery and move repeat orders direct.

Can I message delivery app customers to order direct?

Not using data the app gave you, which the platforms' terms generally prohibit. You can put your own ordering link and QR code in every bag, on your Google listing and on social media, so the customer finds your channel on their own. Once they order on your channel and opt in, you can message them about offers from your own number, in moderation, with an easy way to stop.

How do tiffin subscriptions work on a direct channel?

Sell a monthly weekday plan at a discount to the single-meal price, collect it by card before the month starts, run a four-week menu rotation scheduled by weekday, and set a skip cut-off the evening before with a credit carried forward. Deliver to pickup points such as a leasing office or an Indian grocery store rather than to doors, with door delivery at a small premium. A WhatsApp bot and store page handle the plan, the skips and the payment without an app.

Does Tezzu work for restaurants in the USA?

Yes. US merchants get the same WhatsApp ordering bot, store page, dashboard and staff app, on plans of $49, $69 and $99 a month plus $0.50 per order, with no commission and a 7-day free trial. Customer payments are taken by card through Stripe and confirmed back to the order automatically, menu prices and zone charges are in dollars, and the existing +1 business number is connected through Meta's embedded signup.