When agents stop reading sponsored lists: talabat's 3.4% advertising line
For boards in Bahrain, the UAE, Qatar, Saudi Arabia, Oman and Kuwait, our management consultants start here: talabat earned 3.4% of GMV from advertising and listing fees in Q2 2026, against an adjusted EBITDA margin of 5.0%. That line exists because a person scans a list, and an agent has none to scan.

What a sponsored position is paid to do
A sponsored position sells a person's attention. A customer opens an app, scans a list, and the restaurants and brands that paid for a place near the top are seen first. In Q2 2026 talabat earned 3.4% of GMV from advertising and listing fees. Its adjusted EBITDA margin was 5.0%.
That revenue rests on one behaviour, which is scanning. An agent that places the order on a customer's behalf ranks by price, delivery time and rating. It has no list to scan and no position that can be moved up. The advertiser is paying for a glance that no longer happens.
For the group that sells placements, the question is how much of the line survives. For the group that buys them, the question is what the money was really buying.
Each share of orders placed by agents takes the same share of the ads line with it
0.34 points
On talabat's 3.4% of GMV, agents placing 10% of orders would remove 0.34 points. At 25% the loss is 0.85 points. One full point is gone at about 3 in 10 orders. The arithmetic takes agent orders with no advertising and holds everything else steady.
Source: talabat Q2 2026 results, released 12 August 2026, for the 3.4% and the 5.0%. Scenario: At Par arithmetic.
Who bears the loss
The line is paid by one position and earned by another, so the exposure is not spread evenly across the Control Matrix. For the platform it is revenue that disappears. For the restaurant or brand that has been paying for placement, it is a saving only if the agent still finds and chooses it on price, delivery time and rating.
A Tenant sells mainly through apps and has treated placement as the price of being found. If agents rank on merit, some of that spend was renting a position the business would have held anyway, and some was buying orders it would otherwise have lost. The two look identical on a statement and very different on a board page.
A Carrier earns no advertising line, but it is paid by platforms that do, and a platform with less advertising income will look harder at what it pays for fulfilment. A Merchant with its own channels is less tied to the list and more tied to whether an agent can read its catalogue, prices and delivery promises. A Sovereign that runs its own platform earns this line itself, and will want to decide whether to sell agents a ranking product or reprice visibility before they force the question.
The board's questions
- A Tenant should ask how much of its visibility spend would still be needed if an agent ranked it on price, delivery time and rating alone.
- A Merchant should ask whether its catalogue, prices and delivery promises can be read and compared by an agent without a human in the loop.
- A Sovereign should ask whether to price a ranking product for agents, or to give up part of the sponsored charge on its own terms.
- A Carrier should ask how lower advertising income at a platform would show up in the rate it is paid.
- Every board should ask which of its advertising lines would survive a customer who never sees the list.
From the advertising line to the group's position
We size what is exposed in your margin if agents place a share of orders, set the group's position on platforms and agents through Digital, AI & Agentic Commerce, and test visibility spend against the orders it actually wins through Commerce Strategy & Market Entry. A restaurant group starts with the aggregator margin tool.
Sources
- talabat, Q2 2026 results, released 12 August 2026: advertising and listing fees at 3.4% of GMV; adjusted EBITDA margin of 5.0%.
- Scenario arithmetic at 10% and 25% of orders placed by agents: At Par analysis. Published in Moneyball Finance, 13 October 2026.
Related
- Who owns the customer when the agent holds the memory?If the customer's agent holds the order history, the platform no longer owns the start of the next purchase, and verifiable fulfilment counts for more than visibility.
- When AI chooses for the customer, every position on the matrix movesAgents put paid visibility, the customer relationship and platform terms in question. Tenants, Merchants, Carriers and Sovereigns each face a different answer.
- Digital, AI & Agentic CommerceYour technology plan, set up and connected, including AI in operations and agent-ready commerce.
Bring last month. Leave with a score.
Sit down with Qasim Abbas. One table, your numbers, and a clear read on where you stand against par.