GrabFood vs LINE MAN commission in 2026: which costs your Phuket restaurant less

Published commission ranges put GrabFood at 20% to 30% and LINE MAN at 15% to 25% of each order in Thailand, according to the software vendor guide Klikit. On those ranges LINE MAN is usually the cheaper app, but the ranges overlap, so the rate written in your own two contracts decides the answer for your restaurant. The commission is also only part of what the channel costs: co-funded promotions, in-app advertising and packaging often take more margin than the gap between the two apps.

What each app charges, and what the rate covers

Both apps take a percentage of the menu value of each order, deduct it before they deposit, and negotiate that percentage restaurant by restaurant. Nobody publishes a price list, which is why the only number that matters is the one on your signed agreement.

The market shares come from Khaosod English, reported when foodpanda announced it was leaving Thailand. They are national figures. Your own split between the two apps in Patong, Rawai or Phuket Town may look nothing like them, and that split is in your statements, not in the news.

What the commission covers: the rider, your listing in the app, payment handling, and the app's customer service when an order goes wrong. What it does not cover: your packaging, the discounts you fund, in-app advertising, menu photography, and the staff time an order takes at the pass.

The arithmetic on 200,000 THB of app sales

Here is a worked example with round, invented inputs: 500 orders a month at an average of 400 THB, so 200,000 THB of menu value on one app. Food cost 35%. Packaging 15 THB per order. A 10% discount that you fund on half the orders.

Ten commission points on this example are 20,000 THB a month. If your two negotiated rates are closer, say 25% and 22%, the same table leaves a gap of 6,000 THB, and the promotions you run matter more than the app you favour. Do not take my inputs: run the table on your own.

1. Download the last three monthly statements from each merchant portal.
2. For each app and each month, write down four numbers: gross menu value, commission deducted, promotion cost charged to you, and advertising charged to you.
3. Count the orders and divide, so you have an average order value per app.
4. Weigh and price the packaging of one typical delivery order, box, bag, cutlery, sauce cups, and multiply by the order count.
5. Compare what is left per order on each app. That single number, not the headline rate, tells you which app is cheaper for your menu.

The costs that are not in the commission

Six things quietly move your real cost above the commission line:

• Co-funded promotions. A percentage off, a buy-one-get-one, or a free item is usually funded by you, in full or in part. It shows as a separate deduction, not as commission.
• In-app advertising. Paid placement inside the app is billed on top. It can be worth it on a slow Tuesday in June. It is still a cost of the channel, so count it there.
• Packaging. Delivery packaging is heavier than a dine-in plate. Rice, sauces and anything fried need their own container.
• Free delivery campaigns. When the app asks you to share the delivery fee to appear in a campaign, you are buying visibility with margin.
• Cancelled and refunded orders. A missing item refunded to the customer is food you cooked and lost. Track how often it happens per hundred orders.
• Photography and listing time. Shooting a menu properly and keeping it updated is real staff hours, once at the start and again at every menu change.

One metric settles arguments: take what the app actually deposited in your bank account for the month, divide it by the gross menu value of the orders, and subtract from 100%. That percentage is your true cost of the channel. If the commission on your contract is 25% but the number comes out at 36%, the other eleven points are decisions you made, and most of them you can change this month.

When to raise your app prices, and by how much

Menu prices on an app do not have to match your dine-in prices, and most restaurants that keep delivery profitable set them higher. The arithmetic is simple. If a dish sells at 300 THB in your dining room and the app takes 25%, you keep 225 THB. To keep the same 300 THB you need a menu price of 300 divided by 0.75, which is 400 THB. Add 15 THB of packaging and it becomes 315 divided by 0.75, which is 420 THB.

That is the ceiling, not the recommendation. Covering the commission in full can push you out of the price band customers scroll past. A workable approach:

1. Read your merchant agreement first for any clause about pricing, then decide how far you will go.
2. Raise the dishes that travel well and have the best margin, not the whole menu.
3. Keep one or two visible low-price items so your listing still reads as affordable.
4. Change prices on one app only, and leave the other as it is for two weeks.
5. Measure order count, average basket and review score on both. Keep the rise only if the lost orders cost you less than the recovered margin.

Expect some comments about price. BrightLocal's 2025 consumer review survey, reported by Searchlab, found 89% of consumers expect a business to answer reviews, so answer them calmly and in the language they were written in.

Turning app customers into direct orders

Be honest about what this will not do. A direct ordering channel will not replace the apps. They put you in front of people who have never heard of your restaurant, and in low season from May to October that discovery is worth paying for. The Phuket News reports more than 600,000 restaurant closures in Thailand over three years, with the low season starting earlier. Demand really does collapse: Knight Frank counts January and February as 43% of first-half air arrivals and June as 11%. Keep both apps. Just stop paying a commission on the same customer for the twentieth time.

1. Finish your Google Maps profile. Hours, menu, photos, and an order or message link. For Thai guests and long-stay residents, put your LINE Official Account there. For Western guests, put WhatsApp. Our guide to ranking on Google Maps in Phuket covers the rest of the profile.
2. Open a LINE Official Account. DataReportal's Digital 2026 Thailand report puts LINE at 56 million monthly active users in Thailand, 82.6% of Thai internet users. A broadcast to people who already ordered from you costs nothing per message in commission.
3. Add WhatsApp for foreign guests. The Nation reported business use of WhatsApp in Thailand grew 208% in 2024. Russian and European guests in Rawai or Kata will message a number before they install anything.
4. Check the agreement before you put a flyer in an app bag. Some merchant terms restrict marketing material inside app orders. Your own packaging, your receipt for walk-ins and your table card are not restricted.
5. Be findable in the languages your customers use. Russian visitors read Yandex Maps, where the Phuket restaurant Chekhoff carries 3,508 ratings; Chinese visitors rate Xiaohongshu the most influential channel, at 4.09 out of 5 in a Prince of Songkla University study.
6. Watch AI answers. Search Engine Land reported Google AI Overviews appearing on 387% more restaurant queries in March 2025. Ask ChatGPT and Google where to eat near your street, in each language your customers speak, and note whether you are named.

This is the work Notorioos does for restaurants in Phuket: Google Maps, AI answers, search and the platform each nationality uses, in English, Thai, Russian, Chinese and French, with the inquiries arriving straight at the restaurant. The same commission logic applies to any booking platform, as in our piece on boat tour platform commissions versus direct bookings.

Do this one thing this week

Pull the last three monthly statements from both merchant portals, and work out one number per app: money deposited divided by gross menu value. That is your real commission, and it is almost always higher than the rate you remember signing. Notorioos offers a free Visibility Check that shows where the restaurant appears on Google Maps and in AI answers in five languages, with the three fixes that matter most.