The headline lie
The most common sentence operators say about delivery is some version of: it’s incremental revenue, even at thin margins it’s worth it. The sentence is wrong in a specific way that takes two years to surface.
Delivery isn’t incremental in the sense of free. It’s incremental in the sense of additional — additional kitchen labor, additional packaging, additional refunds, additional manager attention when the platform’s GPS sends someone to the wrong door. Each of those is a real cost, and on a thin-margin order none of them go away just because they don’t appear on the platform’s payout report. The platform shows you a number. The number is gross sales minus their cut. It is not your contribution.
The sheet exists to add the missing lines.
The five lines you need to count
The platform shows you the first three. The last two are on you.
1. Gross sales. What customers paid the platform before any fees. Easy to find on the dashboard.
2. Commission and service fees. The platform’s cut. On most consumer-facing platforms this runs 15–30% of gross, depending on which tier you signed up for and how much marketing the platform is throwing at you. Also on the dashboard.
3. Marketing fees. Sponsored placement, “buy one get one” subsidies you funded, the discount on a customer’s first order — you paid for those. Also on the dashboard, but in a different report than commission. Pull it.
4. Refunds and adjustments. A customer reported a missing item; the platform refunded them out of your account. A driver dropped a soup; you ate it. These show on the dashboard but tend to land in a different month than the order they refer to. Treat them as a real reduction of net payout, not a customer-service blip.
5. Your costs. The big one, and the one the platform doesn’t and won’t show you:
- Food cost — apply your normal food-cost percent to gross. The dish costs the same whether it eats in or rides on a bike.
- Packaging — clamshells, lids, bags, stickers, single-use cutlery. Most operators land in the $0.60–$1.20 per ticket range; if you’re using compostables it can climb past $2.00.
- Incremental kitchen labor — the hardest line to be honest about. The expo who packs delivery, the cook who batches it, the manager who fields the “where is my order” texts. If those people exist because delivery exists, their pay belongs in this line. The most common lie operators tell themselves on this sheet is “the kitchen was going to be staffed anyway, so delivery labor is free.” If the kitchen is at capacity on a Friday night, that lie is exactly backwards — delivery labor is what’s preventing your dine-in turn from happening.
Reading the contribution margin bands
The sheet returns a contribution margin in dollars and as a percent of gross. The percent is the band the channel sits in.
Negative — you’re losing money on every order before fixed costs are layered on. The channel is a subsidy in the literal sense. Action is not optional and is on the next day’s calendar.
0–5% — thin contribution. The channel covers its variable costs but doesn’t pay for fixed costs (rent, insurance, salaried management, marketing’s portion of the calendar). The model is treading water. Either renegotiate, raise menu prices on this channel, or shrink the menu offered through it to higher-margin items only.
5–12% — the typical band for third-party delivery. The channel earns some contribution, but not as much as a dine-in dollar. Watch the trend month-over-month; this is the band where small drift compounds — commission rates creep, packaging vendors raise prices, refund rates climb on a busy holiday weekend.
12%+ — healthy contribution for a third-party channel. Keep an eye on packaging and incremental labor — these are the line items that drift quietly. Most channels that read here have a tight delivery menu (limited SKUs, no items that travel poorly) and an honest labor estimate.
The renegotiation playbook
If your channel reads negative or 0–5%, you have four levers. Pull them in this order.
Lever 1 — Tighten the menu offered through the channel. The fastest contribution-margin improvement is removing the dishes that travel badly. A burger that hits the customer cold is a refund waiting to happen and a one-star review on top. Remove it from the channel menu. Keep the items that travel well and run a high food cost on through (the customer doesn’t see your menu prices side-by-side; you can charge a delivery premium). Most operators see two to four points of margin from this move alone.
Lever 2 — Renegotiate commission. Platforms negotiate. They won’t volunteer it, but a phone call with a sentence like “I’m looking at our channel P&L and the math doesn’t pencil at this rate — what tier can you offer me to keep the volume on the platform” opens a conversation. Most independent operators on the standard tier can negotiate down 2–4 points if they have six months of consistent volume. A 2-point reduction on commission is a 2-point increase on contribution; it stacks with Lever 1.
Lever 3 — Reprice. Some operators run the same menu prices on dine-in and delivery; many run a 10–15% delivery premium. A premium isn’t a tax on the customer — it’s how the channel earns its keep. The customer sees a price; they don’t see the platform’s 28% commission. Charge for the channel’s real cost.
Lever 4 — Turn it off. If the first three don’t move the channel into the green band, the channel doesn’t belong in your operation. Most operators discover one of two platforms is doing the heavy lifting and the other is dragging the average down; turn off the dragger. The customers who actually want delivery from your restaurant will find you on the platform you keep, and the kitchen labor that was supporting both platforms gets reallocated to the one that earns its keep.
When to turn it off
The hardest decision in this sheet is the “turn it off” one. The number reads bad, but you’ve been on the platform for three years and you’re afraid of what the customers will say. The math of that fear is worth doing.
Run the sheet honestly. If contribution margin is negative, every dollar of delivery is a dollar you would have been better off not selling. The customers who order from you regularly through the platform are not net-positive customers — they are net-negative customers, and you have been paying the platform for the privilege of serving them. Turning off the platform is not losing customers; it is stopping the bleeding.
The customers who would have ordered from you anyway will pick up the phone, walk in, or use the second platform you keep. The ones who only existed because the algorithm pushed them there will go elsewhere — which sounds like a loss until you remember each of them was costing you money. The recipe-cost-card on the dishes you keep on dine-in will still tell you what to charge; the prime cost worksheet will get a small lift from the labor you stop allocating to the channel. The sheet won’t make the decision for you, but it will tell you what the decision actually costs.
Open the sheet: Third-Party Channel P&L →
Pull last month’s statement, count all five lines, slide the labor-honesty multiplier. Your numbers never leave the page.