Delivery revenue is not delivery profit.

Is growing delivery revenue good for a restaurant? Only if the channel's contribution — what is left after commission, discounts, packaging, refunds and the operational load — is positive and better than the alternative use of the same kitchen capacity. Gross delivery sales cannot answer that question, because every cost that decides it sits below the line the aggregator dashboard shows you. Plenty of restaurants are growing a delivery number and shrinking a business at the same time, and the P&L only reveals it quarters later.

This article gives you the arithmetic to answer it for your own operation in one afternoon.

Two definitions before anything else

Delivery revenue is the gross value of orders placed through delivery channels — marketplace apps, your own online ordering, phone orders you run out yourself.

Delivery contribution is what those orders add to the business after the costs the channel itself creates: platform commission, payment fees, promotional subsidies and discounts, packaging, delivery-caused refunds and remakes, and the food cost of a menu mix that usually skews differently from the dining room.

Revenue is what the platform reports. Contribution is what pays rent. They move independently, and that is the whole problem.

Where the gap comes from

Six deductions separate the two numbers, and each one is invisible on the order screen:

1. Commission. Marketplace commission is negotiated per operator and varies by market, platform, and plan. Whatever your rate is, it applies to the gross basket — including the portion you discounted.
2. Promotions and discounts. Platform campaigns are usually funded partly or wholly by the restaurant. A "buy one get one" funded by you halves the revenue on the item and leaves the food cost untouched.
3. Packaging. Real, per-order, and rising with order count — exactly the shape of a cost that never appears in a monthly review because no single order makes it visible.
4. Refunds and remakes. A delivery order that arrives wrong is refunded or remade at your cost, for a failure that may have happened after the food left your pass.
5. Menu mix drift. Delivery baskets skew toward mains and away from the high-margin perimeter — beverages, desserts, sides sold by a server who can suggest them. Nobody upsells inside an app unless the menu is engineered to.
6. Operational load. Kitchen capacity, a dedicated packing station at peak, and the service disruption of drivers in the lobby. This one resists precise costing; it should still be named in the review rather than rounded to zero.

A worked contribution view

The figures below are fully illustrative, chosen for arithmetic clarity. They are not benchmarks, not UAE market data, and not a Katalyst client result. Replace every assumption with your own numbers — the structure is the point.

Assume a delivery order with a gross basket of 100 (any currency), a 25% marketplace commission, a restaurant-funded discount of 10 on the basket, packaging at 4 per order, a 30% food cost on the full undiscounted basket, and refunds/remakes averaging 2 per order across the month.

| Line | Amount | |---|---| | Gross basket | 100 | | Restaurant-funded discount | −10 | | Commission (25% of 90) | −22.5 | | Food cost (30% of 100) | −30 | | Packaging | −4 | | Refund/remake average | −2 | | Channel contribution | 31.5 |

The same basket sold in the dining room at full price with no commission and no packaging, and the dine-in beverage attach a server produces, contributes roughly double — before you count what the delivery order did to kitchen pace at 8pm. And if the discount deepens by another 10, contribution drops to roughly 24: the discount came off the top line, but every dirham of it came out of the margin.

None of this makes delivery wrong. It makes delivery a channel with its own P&L, which is how it should be read: some concepts, dayparts and menu lines are genuinely profitable to deliver; others are volume that photographs well.

The Katalyst Delivery Contribution Questions

A Katalyst method — six questions that turn a delivery report into a decision:

1. What is contribution per order, per platform, after every deduction above?
2. Which menu items survive commission — and does the delivery menu sell those, or the dining-room menu shipped as-is?
3. What share of delivery demand is incremental, and what share is your own dining-room or direct guest ordering through a costlier channel?
4. Who owns the guest? A marketplace order usually leaves you no usable relationship — no contact, no preference history, no reactivation path.
5. What does the discount calendar cost per month, and which campaigns did you choose versus default into?
6. What would the same kitchen hour earn if the capacity went to the best alternative — dine-in, pickup, a delivery-only line engineered for margin?

What operators usually find

Three patterns repeat when this arithmetic is actually run. First, one platform is materially better than another for the same food, because commission plans and promotional defaults differ. Second, a handful of items — usually liquid-heavy, discount-exposed or packaging-intensive — contribute nearly nothing and can be re-priced, re-portioned or removed from the delivery menu without touching the dining room. Third, direct ordering is worth building only after the maths shows which orders it should capture; a direct channel that merely re-routes marketplace orders at the cost of your own marketing is not automatically a win either.

Judge the delivery channel the way you would judge a tenant: by what it pays after its costs, not by how often it shows up.

Sources and limitations

Commission rates, promotional structures and fee plans are negotiated per operator and change; no universal percentage is quoted here because none would be honest. The worked example is illustrative arithmetic with every assumption stated, not market data. Where you need your own inputs: your platform statements (commission and promo deductions are itemised), your packaging invoices, and one month of refund logs.

If the picture this builds is uncomfortable, that is the finding. The F&B Growth & Margin Diagnostic runs this arithmetic across dayparts, channels and menu mix as one connected view — part of the F&B Growth & Revenue practice. Related reading: the outlet is full — is it profitable? and private dining is a sales pipeline, not a reservation.