I’m paying for GrabFood ads but orders are not increasing. Why?
The diagnostic order
The most common and most expensive cause. While the restaurant is offline, ads either do not run or lead to a closed listing — and the algorithm remembers the unreliability and cuts organic impressions. USSR Phuket had 3,977 offline minutes a month; once removed, search impressions went from zero to 7,481 a month. Enjoy Healthy Food had a 73% offline rate — brought to 0%, impressions grew from 7,038 to 25,543 a month.
If impressions are there and orders are not, ads are not the problem — the listing is. At Etna Phuket, 39,211 people saw the listing monthly and 182 ordered: 0.5% through-conversion against our 0.9% fleet average. Only 5.2% of viewers even opened the menu. Until that is fixed, every ad dollar buys one more look without an order.
A promo lifts your position and cuts your margin at the same time. What matters is not "how many orders the promo brought" but what is left after the discount, the platform commission and the ad spend. A promo that does not pay back looks like order growth and behaves like a loss.
From 4.8 the algorithm serves impressions more generously, and a customer choosing between two listings looks at the number next to the name. Ads bring them to the listing; the rating decides whether they order. USSR Phuket went from 4.5 to 4.8, Zaytun Ubud from 4.67 to 4.8.
Auto-bidding buys cheap, irrelevant impressions: you pay for people who do not click. At Etna we switched to manual CPO with daily management — CTR went from 2.8% to 5.59%, cost per order from 42 to 29 THB, ROAS from 14.75x to 34.57x. The budget rose only 50% while ads revenue grew 3.4x. At Zaytun Ubud, GoFood campaigns were losing money at 0.25x ROAS — Rp 3.1M spent, Rp 763K returned; after the rebuild, 15.52x on nearly the same budget.
The one mistake
Ads are the only lever that works instantly and shows up in a report, so they get pulled first. The other four steps require daily work and produce no pretty chart by day two. Yet they decide what a purchased impression turns into. Ads on an unprepared listing are payment for the speed at which you lose the customer.
Case studies with the full numbers
Etna Phuket — orders x2.2 on falling traffic, ROAS 14.75x → 34.57x. Zaytun Ubud — GoFood ads from a loss-making 0.25x ROAS to 15.52x. Enjoy Healthy Food — revenue x9.4 in 14 months, offline rate 73% → 0%.
Frequently asked
How do I know if my ads are actually paying off?
Our clients hold 20–27x, higher in individual campaigns. But ROAS in isolation is misleading: it counts ads revenue and ignores your discount, the platform commission, and the share of those orders you would have received anyway. Read ROAS together with through-conversion and post-promo margin.
If I raise my ad budget, will I get more orders?
If the listing does not convert, a bigger budget buys more views without orders — the fastest way to waste money. At Etna we raised the budget 50% only AFTER fixing the listing and the bidding, and got 3.4x the ads revenue.
How soon will I see my orders start growing?
First movement in 2–4 weeks, full ramp-up in 3–6 months. Availability and bidding respond fastest; listing conversion and rating take longer because the algorithm needs history.
My listing is good — do I still need ads at all?
Yes — but as an amplifier, not a substitute. Ads buy impressions; the listing earns the order. In the right order, ads multiply what already works; in the wrong one, they pay for what does not.
What we find on an account like yours
Not hypotheses — what shows up in the first days when we open the dashboard of a restaurant with this complaint.
On average 25% of revenue leaks past the restaurant, and 95% of that loss is switched-off items — not a closed restaurant (3%), not cancellations (2%). We routinely see 40–70 items off at once, and individual dishes stuck in the stop-list for over 2,000 hours. Every paid click in that window sends someone to a menu where half the bestsellers are missing.
That is the line where ads stop paying back: below it the median ROAS is 12.1x, above it 8.6x. 42% of the restaurants in our fleet are already past that line. The symptom is exactly the one owners arrive with: the budget grows, the orders do not.
The Bali medians we check every account against: Rp 250k average check, 10.4x ROAS, ads at 5.6% of revenue, 0.35% cancellations. Put your own ad share of revenue and your own ROAS next to them — that pair shows immediately whether you are buying orders or impressions.
Reviews are bimodal: 51% five-stars, 28% one-stars and only 3% fours — the extremes make the rating. The Bali median is one negative review per 138 orders. Around 80% of the appeals we file with Grab end with the review removed, but usually nobody is filing them. Meanwhile the ad traffic lands on a listing whose number is already down.
Keep reading
Two ways to go from here
Both work. The first costs nothing and does not require us.
The method and the norms — open and free
The five stages we run on every account, published in full. Alongside them, market norms from 96 restaurants so you have something to compare your numbers against.
An audit of your listing
Paste your restaurant’s Grab link and the report comes back in a couple of minutes: menu and search, photo coverage, reviews, prices against the neighbours. Free, no strings; after that it is 10% of delivery revenue with no upfront payment.
Diagnose my listingor message us directly →