Does my restaurant need a website if it's already on Zomato and Swiggy?

If you run a restaurant around Udupi, you know the payout statement. The orders were good this month, and then the deductions arrive: commission, GST on the commission, a discount you don't remember agreeing to, an ad you didn't know was running.

Some of those orders came from people who'd never heard of you. Plenty came from your regulars, who would have ordered anyway.

So it's a fair question: would a website of your own win some of that back? Here's the honest answer, starting with what the apps do better than any website I could build you.

What Zomato and Swiggy are genuinely good at

• Delivery. They have the riders. A website doesn't come with a fleet, and most restaurants don't want to run one.
• Hungry strangers. A student in Manipal who wants dinner in thirty minutes opens an app, not a search engine. Being listed is how they find you.
• Payments and refunds. Card, UPI, cancellations and complaints are handled somewhere other than your counter.

That last point cuts both ways, as we'll see. But the first two settle something up front: a website will not replace the apps for delivery to people who don't know you yet. Keep your listings.

What it costs you, and why it keeps going up

There's no single commission rate. It depends on your city, your cuisine and what you negotiated. The Week reported in August 2026 that the typical rate is 24 to 25 per cent. And the commission isn't the whole bill. GST is charged on it, and so are the extras: discounts the platform runs, ads, deductions after a complaint.

Those extras are what pushed Bengaluru's restaurants to the edge this year. In July, restaurant associations there threatened to stop taking orders from both apps from 15 August. Their complaints weren't only the commission rate. They were promotions and discounts run without written consent, money deducted automatically after a customer complaint, and monthly statements that don't itemise the deductions. Swiggy agreed to refund promotional charges levied without consent, and the deadline was pushed back.

Work it out with your own numbers, not anyone's average, mine included. Take last month's statement and add up everything that isn't food: commission, GST on it, discounts you funded, ads, deductions. Divide by the order value. If it comes to 25 per cent, a ₹400 order hands over ₹100.

Then ask the question that matters: how many of those orders were from people who already knew you?

The commission on customers who were already yours

A commission makes sense when the app finds you a customer you'd never have met. It's charged just the same when it didn't.

Think about who actually orders you through the app:

• The office down the road that orders lunch from you every Friday, through the app, because that's where the Order button is
• The family who eats at your place on Sundays, and orders through the app on weeknights
• The person who walked past, liked the look of it, and looked you up later

Each of them chose you first. The app only took the order, and it charged the same commission as on a stranger.

Not all of those orders can come back to you. Delivery is still delivery, and someone has to make it. But some of them can: pick-up orders, orders from close by that your own boy can run, and the one category that should never have gone through an app at all.

The orders that should never go through an app

Party orders. Catering for a function. Fifty idlis and a vessel of sambar for an office pooja. A naming ceremony at home.

These are your biggest orders, the apps aren't built for them, and the people placing them want to talk to a person first. They search your name, look for a phone number, and call. If what they find is a Zomato page, the conversation never starts.

A page on your own site that says "we do catering, here's roughly what it costs, here's the number" is often the most valuable page a restaurant website has.

What people search before they come to you

Start typing your restaurant's name into Google and look at what it suggests: menu, timings, phone number, "open now". People want to know whether you're open, what you serve and how to reach you, before they walk in or call.

What they find decides what happens next:

1. Only your app listings. They see app prices, which are often higher than your counter, and order through the app or not at all.
2. A Google Maps listing with old timings and no menu. They wonder whether you're open on Mondays. Some of them go somewhere else.
3. Your own site. Today's timings, the full menu with prices, the WhatsApp number. They come in, or they order direct.

In Udupi there's an extra twist: "Udupi" is on restaurant boards across the country. A search for an Udupi restaurant returns places in Bengaluru, Mumbai and Chennai. Your full name, your area and your own site are how Google tells your place near the temple apart from a namesake 400 km away.

Your restaurant's name is a search you can actually win, because the results competing for it are mostly your own listings. Put the name and the area in your site's title, link the site from your Google Maps listing, and once Google has indexed it, it usually shows up alongside them. It's the same moment an Instagram follower Googles a business before buying, with a meal instead of a product.

What the site actually needs

Not twenty pages, and not your own delivery app.

1. The menu as text, with prices. Not a photo of the menu card and not a PDF. Text loads fast on a phone, Google can read it, and nobody has to pinch and zoom to find the dosa section.
2. Timings, including the exceptions. The weekly holiday, festival closures, when the kitchen stops taking orders.
3. Exactly where you are. A map, the landmark people actually use, and whether there's parking.
4. Who it suits. Pure veg, Jain options, family seating, room for a group of twenty, AC section.
5. Catering and party orders. What you do, the minimum quantity, how much notice you need, the number to call.
6. A WhatsApp button, with your number written out as well. Some people will want to copy it or call.

And it has to load fast. Someone deciding where to eat in the next ten minutes won't wait for a slow page, and a slow site loses exactly the customer who was ready to come.

Turn app customers into direct customers, carefully

• At the counter and on the table. A small card with your website and WhatsApp number, and a QR code on the table that opens the menu page.
• In the bag, only if your agreement allows it. Some platform contracts restrict what you can put in a delivery. Read yours before you print anything; your listing is worth more than one direct order.
• Wherever your name already appears. Your Google Maps listing, your Instagram bio, your WhatsApp Business profile.
• Price carefully. The Competition Commission of India's investigators found that both apps' contracts included price-parity clauses, which stop restaurants offering lower prices on their own channels. That case hasn't been finally decided, so check your contract before charging less on your own site. You can always offer what the app can't: a free side with pick-up orders, a regulars' card, the table by the window.

When the apps are enough on their own

This is true more often than a web designer would like to admit.

If you're a cloud kitchen with no dine-in and no catering, almost every customer really is a stranger the app found for you. If nobody can answer the phone or WhatsApp at lunch rush, direct orders will go unanswered, and an unanswered order is worse than an app order. And if the apps keep you busy and the margin still works, that's a business decision, not a mistake.

Questions people actually ask

Will a website get me orders from Google?

From people searching for your restaurant by name, and for your menu or timings: yes, soon after Google indexes it. From people searching "best restaurant in udupi": rarely, and not for a long time. Those results go to directories, food blogs and the apps themselves, and ranking for competitive searches takes months even when everything is done right.

Do I need my own online ordering system?

Not to start. A menu page and a WhatsApp number cover pick-up orders, nearby deliveries and catering enquiries for most restaurants. Add an ordering system later if the messages become more than you can keep up with.

Should I leave Zomato and Swiggy?

For most restaurants, no. They're still where hungry strangers look. The point of your own site is to stop paying commission on the customers who were coming to you anyway, not to give up the ones you'd never have met.

What about ONDC and the newer apps?

There are more options than two now. Apps on the government-backed ONDC network, Flipkart's new food delivery service and Rapido's Ownly have all pitched themselves to restaurants on lower or flat fees; The Week put Flipkart's rate at 10 per cent. They're worth a look, and worth comparing with your own numbers. But they're still someone else's platform, and the same question applies to them: who is the customer, theirs or yours?

What does a site like this cost?

What a small business website costs in India has its own post. For a restaurant, the useful comparison is that price against one year of commission on customers who already knew you.

Who should own the domain?

You: registered in your name, on your account, even if someone else builds the site. Ten minutes will tell you whether that's true today.

The apps are a good way to reach people who've never eaten your food. They're an expensive way to keep the ones who have.

If you run a restaurant around Udupi and want an honest opinion on whether a website would pay for itself, book a free call. If the answer is no, I'll tell you.