Delivery mistakes that kill your margin

In many restaurants, delivery starts as an “extra” and ends up becoming a meaningful share of sales. The problem appears at month-end, when results feel contradictory: “we sold more… but we didn’t earn more.” Sometimes it’s worse: volume grows while profit drops.

This doesn’t happen because delivery is “bad.” It happens because delivery is a different channel, with a different economic structure. Costs change, customer behavior changes, the sales mix changes, and operations change. If you manage delivery as if it were “dine-in with a bag,” friction will eat the margin.

This post is a didactic guide to where money disappears in delivery—and how to cost it in a simple way so you can decide what to sell, how to price, which conditions to set (minimums, bundles), and when to say “this isn’t worth it.”

1) The core idea: delivery isn’t a service, it’s a channel with its own P&L

Dine-in has one logic. Delivery adds a second layer:

• Platform fees (marketplace commissions)
• Packaging costs
• More incidents (refunds, remakes, poor arrival quality)
• More “operational waste” (re-cooks, long times, queues)
• A different sales mix (different items sell than in dine-in)
• Different kitchen pressure (peaks that don’t match reservations)

If you don’t separate the channel, delivery distorts your Food Cost, Prime Cost, and your sense of control.

Best practice is simple: cost and analyze delivery separately.

2) The minimum formula: does an order actually make money?

A simple formula explains most of delivery economics:

Margin per order = Net order revenue − (Food & beverage cost + Packaging + Incidents + Additional operating cost)

Here’s the didactic point: in dine-in you often think “price minus dish cost.” In delivery, if you stop there, you’re missing half the map.

2.1) What “net revenue” means in delivery

It’s not what the app shows as “sales.” It’s what remains after:

• platform commission (percentage)
• any fixed per-order fee
• payment processing fees (depending on model)
• promotions/discounts you absorb (not the platform)

Rule #1: work with net revenue, not gross.

3) A realistic example (with numbers): where margin goes

Imagine a typical order (one dish + one drink/add-on) with a ticket of €24.00.

Assume:

• Platform fee: 30%
• Total packaging (containers + bag + cutlery): €0.90
• True food cost (costing): €7.20
• Average incidents (refunds/remakes/errors), allocated per order: €0.60
• Additional operating cost (extra time/queues/duplication), conservative allocation: €0.50

Step 1: net revenue

30% of €24.00 = €7.20
Net revenue = 24.00 − 7.20 = €16.80

Step 2: subtract real delivery costs

Costs = 7.20 + 0.90 + 0.60 + 0.50 = €9.20

Step 3: margin per order

Margin = 16.80 − 9.20 = €7.60

Key learning: if you were only looking at “24 − 7.20 = 16.80,” you were massively overestimating margin.

And this is before a common second-order effect: delivery can increase workload at critical moments and reduce dine-in margin (delays, mistakes). That’s why in consulting we look at direct delivery margin and its impact on the full operation.

4) The 9 most common delivery mistakes (and how to fix them)

Mistake 1: same pricing as dine-in

Fix: price by channel or redesign product + raise average ticket via bundles.

Mistake 2: not allocating packaging

Fix: packaging as a per-order or per-item line cost.

Mistake 3: ignoring incidents

Fix: monthly incident cost / number of orders = average cost per order.

Mistake 4: selling “delivery enemies”

Fix: a travel-proof delivery menu.

Mistake 5: letting delivery peaks collapse dine-in

Fix: production windows, dedicated station in peaks, or capacity limits.

Mistake 6: not managing sales mix

Fix: delivery-specific menu engineering (push 4–6 travel-proof high-margin items).

Mistake 7: discounts without a model

Fix: define a minimum margin per order; don’t cross it; use bundles/upsells.

Mistake 8: no channel-specific costing

Fix: dine-in costing + delivery costing for top sellers.

Mistake 9: not considering “contribution per kitchen minute”

Fix: margin plus production simplicity (mise en place, steps, ticket time).

5) How to cost delivery without getting lost: the 3-layer method

Layer 1: base dish cost.
Layer 2: channel costs (fees + packaging + incidents).
Layer 3: operating allocation (if delivery adds real labor/time).

6) Decisions that usually improve margin without killing sales

Bundles to raise ticket, minimum order, shorter travel-proof menu, standardized assembly/labels, and separating capacity during peaks.

7) When delivery is not worth it

Low margin per order that won’t improve, high incidents, dine-in damage during peak hours, discount dependency, and no capacity control.

8) Conclusion

Profitable delivery is designed and controlled. Improvements usually come more from mix, portions, packaging, and capacity than from wishing fees were lower.