Your own website vs Swiggy and Zomato — the real commission maths

I spent five years in a hotel kitchen in Chhatrapati Sambhajinagar before I started building websites. So when a restaurant owner asks me about an ordering page, I am not thinking about design. I am thinking about the ticket printer.

Here is what nobody outside a kitchen knows: an aggregator order and a direct order look identical at the pass. Same docket, same dish, same hands, same gas. The difference is invisible where the food is made and very visible at the end of the month.

The conclusion up front, because articles that hide it are usually selling something: your own website does not replace Swiggy and Zomato. It does a different job. Anyone telling you to delist and "save the commission" has never watched the order count in week two.

First, the honest part about the numbers

Neither platform publishes a rate card for restaurants. Terms are negotiated per outlet and vary by city, cuisine, listing age, and how hard the person before you argued.

Industry coverage through 2026 puts base commission in the 18–28% band. Treat that as a reported range, not a fact about your restaurant — the only documents that tell you your real number are your contract and your payout statement. And commission is not the whole deduction: you may also be carrying discounts you fund yourself (that "50% OFF" badge is usually your money, and the cost owners most often forget), ad spend to stay visible, payment charges, and packaging you never fully recover.

An example plate, worked out

The numbers below are an illustration I made up to show the shape of the problem — not any real restaurant's accounts. Your figures will differ; the point is the method. Say a dish you sell at ₹350 at the table is listed at ₹400 on the app, because you padded the price to absorb commission. Assume a 20% restaurant-funded discount, 25% commission, 2% payment charges.

Through the app

• ₹400 less your own 20% discount → order value ₹320
• Commission 25% −₹80, payment charges −₹6 → you receive about ₹234
• Food cost −₹115; packaging — container, lid, foil, bag, spoon −₹25
• Left for gas, labour, rent, electricity: about ₹94

The same plate ordered directly

• Your own price ₹350, no padding; payment charges −₹7
• Rider or third-party delivery −₹50; food and packaging −₹140
• Left over: about ₹153

Roughly ₹59 more on the plate — and the customer paid ₹50 less. That is the whole argument for direct ordering.

But note what I quietly assumed: that the order arrived at all. On the app it arrives because fifty million people are looking at the app. On your page it arrives only if somebody already knew your name. That is the honest limit of everything here.

What the spreadsheet does not show you

Commission is the cost everyone sees. These are the ones I learned standing at the range — and they are why two restaurants on the same rate end up in very different places.

A flash sale is a kitchen event, not a marketing event

When a campaign goes live, the printer does not give you eleven orders over an hour. It gives you fourteen tickets in six minutes. The kitchen does not scale — same four people, same four stations. So the tickets in the middle of that rush leave at your worst quality of the day, for new price-led customers most likely to rate you. You paid for the discount, then paid again in ratings.

Not hypothetical: in the same letter, Eternal says it cut the free-delivery minimum to ₹99 from ₹199 for Gold members and curated meals under ₹250 — and that a falling average order value is "a natural and intended consequence of this strategy". Read that as more tickets, smaller tickets, same number of hands. Their economics improve on volume; yours do not automatically follow.

Some dishes should never have been listed

A biryani travels. A dosa does not. Gravy in a thin container arrives as a disaster and a ₹40 refund. Every kitchen has three or four items that are excellent at the table and embarrassing after twenty minutes in a bag — usually listed because somebody uploaded the whole menu card at once. And the rider who took two other drops on the way is not your employee, but "food was cold" lands on your rating.

Nobody at your restaurant has the phone number

If I could say one thing, this is it. A man orders from you forty times a year — your best customer — and you cannot contact him. No number, no name, no way to mention the Sunday thali. When someone is unhappy with a direct order he messages you and you fix it with a sweet next time; on the app he tells four thousand strangers, and at low volumes one bad rating drags your visibility.

Where the aggregators genuinely earn their cut

I would not trust this article if it only ran one way, so — the other side:

• Discovery. Someone who has never heard of you will never type your name into Google. The app puts you in front of him; no website you own does that in year one.
• Logistics at 11pm in the rain. Running riders is a whole business with its own hiring and attrition, and owners badly underestimate it. Payments, refunds and cancellations are handled too.
• The demand is real and growing. Eternal reported food delivery net order value up 18.8% year on year in the quarter ending March 2026. Walking away from that is not clever, it is a smaller business.

So the question is never "which one". It is: which orders should be paying 25%, and which are paying it for no reason? A stranger discovering you is worth a commission. A regular who searched for you by name is paying a finder's fee on a customer you found yourself.

The rule that used to block all of this — and what changed

For years there was a practical reason restaurants could not just price lower on their own channel: price parity clauses. When the Competition Commission of India ordered its investigation in April 2022, it noted the agreements appeared to stop restaurant partners keeping lower prices or higher discounts on their own channels or any other aggregator, as reported by PTI. Plainly: you were not free to make your own website cheaper.

That has moved. Business Standard reported in July 2026 that the CCI Director General's investigation concluded three categories of Zomato's contractual arrangements — exclusivity conditions, minimum business guarantees and wide price-parity clauses — contravened Section 3(4) read with Section 3(1) of the Competition Act, 2002. The same report quotes a company source saying Zomato "does not have exclusivity built into its standard agreements and removed price parity requirements in April 2026 or earlier in 2026". The NRAI has asked the CCI to stop those clauses being revived while the case runs.

So what is your own site actually for?

Not for beating the apps at delivery. For three jobs they are structurally bad at:

• Catching people who searched for you by name. Your highest-intent traffic — and today it lands on an aggregator listing, where you pay commission on a customer you already earned.
• Big orders that never come through an app. Party orders, catering, bulk thalis, office tiffin, festival trays. Nobody books a fifty-plate Ganpati order on a delivery app — they call. If your page does not make that number obvious, they call whoever's does.
• Holding your own customer list, and being the one place where your price is your price.

What goes on the page — menu as real text, honest photos, timings, buttons — is in what a restaurant website should have.

Start much smaller than you think

1. Fix your Google Business Profile first. Free, and in this trade it brings more new customers than the website will, because "restaurant near me" runs through it — setup here.
2. One page: menu, real prices, WhatsApp order button. Not an app, not a cart.
3. Take payment on UPI — the QR is already at your counter. Offer pickup at a small discount; it carries no delivery cost at all.
4. Put the link everywhere — bill, board outside, packaging sticker, Google profile. Your existing customers are the entire audience.
5. Keep both channels running and compare payout per plate after three months, from your real statements — not from anybody's blog example, mine included.

Not a big project — roughly the commission on a few busy weekends. Pricing logic is in what a website costs in India; local rates are on my Chhatrapati Sambhajinagar page.

Frequently asked questions

Should I leave Swiggy and Zomato if I build my own website?

No. Almost every restaurant that delists to save commission just loses those orders, because the demand was the platform's. Run both: aggregators for discovery and strangers, your own page to stop paying a finder's fee on people who already know your name.

How much commission do Swiggy and Zomato actually charge?

Neither publishes a rate card, and terms are negotiated per outlet. Reported 2026 ranges sit around 18–28% base commission, but the effective deduction is higher once self-funded discounts, ads and payment charges are counted. Work out your own: take one month's payout statement and divide what reached your bank by the total order value for that month. That number is usually a surprise.

Can I keep lower prices on my own website than on the apps?

The direction has moved in restaurants' favour — wide price-parity clauses were among the arrangements the CCI Director General's investigation found in contravention, and Zomato is reported to have removed them during 2026. But the case is pending and terms vary by outlet, so check your own agreement first. Even where parity applies, you can still offer what the app cannot carry: pickup discounts, family packs, party rates.

Do I need a payment gateway and my own delivery riders?

Not to start — this is where owners overspend. A WhatsApp button, your existing UPI QR, and either pickup or whoever already delivers for you covers the first months. Hiring riders is a separate business, rarely worth it for one outlet.

Will my own website really get orders?

Not from strangers, and not quickly — it will not out-rank a delivery app for "biryani near me". What it reliably converts is people who already know you: someone who searched your name, saw your board, or ate at your table last week. Smaller pool, but each order is worth meaningfully more per plate and those customers stay yours.

If you run a restaurant, dhaba or cloud kitchen here and want a straight opinion, send me your menu and your current commission rate. If your numbers say stay on the apps and do nothing, I will tell you that.