GrabFood Account Optimization for Restaurants in Thailand

In this guide

1. The Thai platform map
2. What the platform actually takes
3. The 9 percent commission
4. Promotions and the free item trap
5. The Thai year
6. Five checks that move an account
7. Questions
8. Getting it done

Thailand is one of the markets we read most. That matters less for the tactics, which travel, than for the four things below: which platforms are worth an account here, what actually separates a Thai merchant's gross from what lands in the bank, the commission round most operators miss, and when in the year each move is worth making.

The Thai platform map, September 2026

The market consolidated hard in 2025. foodpanda closed its Thai operation on 23 May 2025 after thirteen years without a profit, and handed its restaurant and customer base to Robinhood. What is left is a duopoly with one real third player.

Thailand's delivery GMV rose from about 4.2 billion US dollars in 2024 to about 5.1 billion in 2025. The growth is real, and it is concentrating: Grab and LINE MAN together went from 86 to nearly 88 percent of the market in a single year.

What this means for you. A Thai package is GrabFood and LINE MAN. Each is a separate account with its own menu, its own reviews, its own campaigns and its own ranking. They drift apart the moment only one of them is maintained, and the drift always takes the same shape: the second account keeps the prices and the photos from the day it was built. We price each platform separately for that reason. Two platforms in one order take 5 percent off.

What the platform actually takes

Most operators know their headline commission and nothing else.

Commission terms on GrabFood are negotiated per merchant and are not published. No honest article can tell you your rate, and any article that quotes one as a benchmark is quoting somebody else's contract. What an article can do is name every line that sits between the order and your bank, tell you where each one is displayed, and give you the arithmetic to work out your own numbers in about ten minutes.

There are five deductions. Only the first is in your contract.

Work out your own two numbers

Neither of these is displayed anywhere in the merchant app. Both take minutes.

Your real commission rate. Open four completed orders of clearly different sizes in Finance, transactions, payment detail. For each one:

commission deducted ÷ (order value − merchant funded discount) = your real rate

If all four agree, that is your rate and you can plan against it. If they do not agree, or if any of them disagrees with the percentage the screen prints beside the commission line, you have found something worth raising with merchant support before you build a price list on top of it.

Your all in take rate. From the Finance summary for a full month:

earnings for the period ÷ gross sales for the period = what you actually keep

Run it every month. It is the only figure that tells you whether the account is getting better or worse, it moves when your promotion mix moves even though your contract has not changed, and almost no operator has it.

The line that moves without anyone touching your contract

Step up commission is the least understood deduction on a Thai Grab account. Grab's own explanation of it, shown inside the merchant app, reads:

A fee for Grab campaigns you've joined (like Storewide Discount or Lower Delivery Fee) to boost your visibility and sales.

That is the whole mechanic. It is not fixed, it is not negotiated, and it is not an error. It rises when you join campaigns and falls when you leave them. Which means that when you evaluate a campaign, its cost is never just the discount you fund. It is the discount, plus the marketing success fee, plus the movement in this line. Compare your take rate for the month before you joined against the month after. That difference is the campaign's true price.

More on the mechanics in our guide to Grab commission fees and what to do about the part you control.

The 9 percent commission most Thai restaurants never claim

This one is Thailand only, and it is the largest single lever on the whole fee stack.

When the Thai government runs a co payment stimulus scheme, Grab discounts merchant commission on orders placed under it. In the Kon La Krueng Plus round in late 2025, the rate was 7 percent for restaurants that registered on the opening day and 9 percent for everyone who joined afterwards. In the Thais Help Thais Plus 60/40 round in June 2026, the rate was 9 percent, alongside campaign and voucher support worth up to 20,000 baht per store and access to a cash credit line.

Set against a normal negotiated rate, that is roughly half your commission on every qualifying order, for the length of the scheme.

How registration works. You join the government scheme through the Thung Ngern app, select Grab as your platform, copy your store code out of GrabMerchant and paste it in, then confirm by OTP. It takes minutes.

Why operators miss it. The terms change every round, the announcement window is short, and the best rate has now twice gone to day one registrants. The thing that actually costs money is finding out late. Restaurants that are already registered simply opt in. Restaurants that are not spend the first week of the scheme filing paperwork while their neighbours sell at 9 percent.

What to do now. Confirm your store is eligible and your Thung Ngern registration is current, before the next round is announced rather than after.

Promotions, and the trap inside the free item format

GrabFood Thailand gives you five self service promotion formats: a bundle, a percentage discount, a specific price, a delivery fee discount, and a free menu item. They behave differently and they are not interchangeable.

The free menu item format is the strongest of the five for building basket value, and it carries the sharpest trap in the entire tool.

The mechanics. One free item and one minimum order value per campaign. The free item must be a priced row on your menu, and it cannot carry option groups, so a dish with a size choice or a sauce choice is not eligible until you create a plain version of it. Grab's own tool suggests setting the minimum at twice the item's price.

The trap. The format has no tiers, and two campaigns do not behave like tiers. If you run a free side at 450 baht and a different free item at 650 baht, an order of 650 receives both. There is no setting anywhere in the interface that switches the lower gift off once the higher threshold is cleared. Operators build these as a ladder, expecting the customer to trade up from one gift to the other. What actually happens is that the largest baskets, the ones carrying the best margin, give away two items instead of one, and the giveaway grows silently with order size.

The rule. One free item offer, at one threshold, at a time. Choose the threshold by what it does to the orders sitting just below it, not by what it costs at the top.

And on co funded campaigns. A storewide discount that Grab co funds looks free and is not. The campaign screen states the marketing success fee before you join, as a share of the discount value. Read it there, add the movement in step up commission, and you have the real price. Co funded is cheap. It is not free, and the cost does not appear on the campaign screen you approved.

The Thai year moves demand more than any tactic

10 to 18 October 2026: the vegetarian festival

Nine days when a large part of the country changes what it eats. Kin jay is nationwide, largest in Phuket, and it is the most predictable demand event on the Thai calendar for any kitchen that can serve plant based food.

It is won in September. A jay section that is live, photographed, correctly named in Thai and already accumulating orders on day one will rank for the whole festival. A jay section built on day three ranks on day eight, when it is nearly over. If you are reading this in early September 2026, you have about four weeks, and four weeks is the right amount.

November to April: high season, if your customers are tourists

Cooler weather, tourists, more orders, and more competitors bidding for the same placement. This is when ad budgets step up, one increment at a time against a return floor, rather than doubling and hoping.

The distinction most guides miss is that high season is a tourist phenomenon, not a Thai one. In Phuket, Samui, Krabi and Chiang Mai the swing is enormous. In residential Bangkok it is mild, and a restaurant serving offices and condominiums should not budget as though it were on a beach. Read your own months before assuming the national pattern applies to you.

13 to 15 April: Songkran

Bangkok empties. Provincial towns fill. If you are in the city, your volume drops for a week and no amount of ad spend recovers it, so the correct move is to spend less, not more. If you are upcountry, it is one of the best weeks of your year. Adjust hours and budget by where you are, not by the national calendar.

May to October: rain, and the rebuild window

The wet months lift delivery demand sharply on the worst weather days and flatten it the rest of the time. Net, it is the trough. It is also the right window to rebuild: this is when prices get tested, dead menu rows come off, photographs get reshot and the listing gets restructured, because a mistake costs less now than the same mistake will cost in December.

Five checks that move a Thai account first

In the order we run them.

1. Both name fields, Thai and English. Thai customers search in Thai. A menu named only in English is invisible to a large share of in app search in its own market. This is the cheapest fix on the list and one of the most common gaps we find on foreign owned restaurants.
2. Photo coverage on every row, not most rows. An unphotographed item converts far below a photographed one and drags down the category it sits in. Coverage first, quality second.
3. Category order and menu depth. Your first two categories carry most of the traffic. A menu that opens on drinks is spending its best placement on its worst margin. See how the Grab algorithm decides whether customers ever see you.
4. Prep time against your delay rate. Late orders are a ranking input, not only a service problem. Open the Operations tab in Insights and read the spikes rather than the average, because an acceptable monthly figure often hides three bad shifts. Related: reducing cancellations on Grab.
5. Your full review history, including the removed ones. The merchant view shows reviews that no longer display to customers. They are frequently the most useful thing on the whole account, because they name specific, checkable failures that are usually still happening. See how to get Grab reviews.

If none of this moves your numbers, the problem is upstream of the tactics. Start with low orders on Grab.

Frequently asked questions

Getting it done, on your account

Three ways in, each priced per platform. The rebuild contains the audit, so you buy one of the two, never both.

After the work, Account Care keeps the account read and adjusted month by month, from $129. Prefer to run it yourself? The books start at $29.