Delivery profitability checklist: each box with its 'done' criterion (unaudited vs audited) 2026

Box 1: I have the effective commission per platform as a number

This box is checked only with a calculated number, not the contract commission. The proof: take the last 90 days per platform, add commission paid, packaging per order, and refunds on incomplete orders, and divide by gross channel sales. Platforms charge between 15% and 30%: on a $25 order at 25% that is $6.25, but with $1.10 packaging and a 4% refund rate, the channel takes $8.35 — 33%. If you think you pay 25% but do not have it documented per app, the box stays empty. Masterestaurant lets no group move forward without this figure, because costing on the 25% contract leaves 11 points uncovered. Diego F. Parra says it plainly: what is not a number is not measured, it is assumed, and in delivery assumptions cost $1.80 to $2.50 of cash per order without the owner noticing.

Box 2: my app price is adjusted per channel, not copied from dine-in

This box is checked only with a 15-22% reprice on the dishes that can absorb it, not with the intention of 'having it under control.' The proof is margin: a dish with 30% food cost leaves 70% in the dining room, but on the app at 33% commission it leaves barely 37% before labor, so the app price must rise to keep equivalent margin. The box requires the per-dish profitability analysis that justifies each reprice, and that you did not blindly raise price-sensitive dishes. Here per-channel AI pricing helps decide which dish can take a +24%. Checking this box correctly moves the channel margin from -$1.80 to $4.10 per $25 order. The mistake I see over and over is owners leaving this box empty out of fear of raising prices, handing $5.90 of swing per order to the platform every day that passes.

Box 3: I have a delivery P&L separate from the dining room

This is the most revealing box and the one most owners check wrongly. It closes only with a delivery-channel income statement apart from the dining room: gross sales, total effective commission, packaging, specific food cost, contribution margin per order, and break-even in number of orders. Without that document, the box stays empty even if you 'feel' delivery is doing well. When the channel lives hidden in the dining-room P&L, a delivery losing 7 margin points disappears from the radar and you believe all is well because total cash rises with app volume. In 70% of the groups Masterestaurant audits, delivery was destroying margin without being measured in isolation. The box is checked by calculating how many orders the channel needs to cover the payroll, rent, and riders it consumes: if it does 3,000 orders at $4.10 contribution margin, it contributes $12,300 monthly to fixed costs.

Box 3: I have a delivery P&L separate from the dining room — in practice

That number, not the feeling, closes the box. This box verifies you respect the hard costing rule, and it is checked only with the layers properly separated. Platform commission is subtracted when costing the delivery channel because it is a direct variable cost of the order; food cost maximum is 32% — never recommended — and measures ingredients only; payroll, rent, and utilities are NOT charged to the plate or the order, they belong to the break-even point of the business, calculated separately. If you lump commission and payroll into the same costing, the box stays empty because you are costing wrong: you end up raising the wrong app prices or subsidizing the channel with the dining room without seeing it. Confusing these two layers is the root of half the delivery errors I find in consulting. Masterestaurant always separates: food cost measures ingredients, commission to the channel, payroll and rent to the monthly break-even.

Box 4: I separate commission from payroll in my costing

The proof is a costing sheet where each cost sits in its correct layer, not everything mixed into a single per-dish number. This box is not checked for 'having a website': it is checked when a gateway at 4-6% is operating and the direct channel already migrates real volume from the app. A direct channel charges 4-6% versus the platforms' 25%, so every migrated order recovers nearly 20 points of commission. It is not about abandoning the apps, which bring reach and new customers; it is about turning the recurring customer into a direct one with a simple incentive, like 10% off their second order through your channel. Even giving away that 10%, you still win against the app's 25%. The box closes when the direct channel exceeds 15-22% of delivery, measured month over month. Masterestaurant groups reach 22% in 6 months; on 3,000 orders per month that returns $9,000 to $13,000 monthly.

Box 5: I have a direct channel operating at 4-6%

Checking this box 'because I have a website' while the direct channel sits at 3% is self-deception: the proof is migrated volume, not the channel's existence. This box turns profitable delivery into a habit, not a one-time adjustment. It is checked when you review monthly, with the formality of a board KPI, what percentage of delivery goes through the expensive app at 25% and what percentage through the direct channel at 5%, with the margin per order of each and its month-over-month evolution. The mistake I see over and over is owners who build the direct channel, let it stall at 8%, and never push it because nobody reviews the mix. The box requires an owner, a monthly cadence, and a target number: the Masterestaurant goal is to take the direct channel from 0% to 22% in two quarters and sustain it. Without this box checked, all the work of the earlier boxes degrades: the mix tilts back toward the expensive app and commission eats the cash again.

Box 6: I review channel mix monthly as a board KPI

The proof is the monthly review minutes with the mix figure, not the good intention of 'keeping an eye on it.' This box takes delivery from profitable to optimized, and it is checked when you apply dynamic per-channel pricing with artificial intelligence. AI cross-references demand by hour, commission by platform, and price elasticity by dish to decide which app prices to raise without scaring off orders: it is not raising everything 18% blindly, it is raising the dish whose demand can take it by 24% and leaving the price-sensitive one untouched. Groups Masterestaurant supports with per-channel pricing raise the average app ticket 11% while losing under 3% of orders, a trade that almost always leaves net positive cash. The box is not mandatory to start — first you close measurement and costing — but it is to squeeze the channel. The proof is an app ticket measured before and after dynamic pricing, with the order variation documented.

Box 7: I use AI for dynamic per-channel pricing

Without AI, the owner sets one price for every channel and gives away margin on dishes that would have absorbed a higher price without losing demand. The eighth box is the closing one: count how many of the previous seven you checked with real proof, not with feeling. Checking fewer than 6 of 8 with evidence means your delivery almost certainly loses $1.80 to $2.50 per order without you seeing it, hidden in the dining-room P&L. Boxes 1 to 4 — effective commission, per-channel reprice, separate P&L, and cost layers — are the minimum to stop losing cash; boxes 5 to 8 take the channel from profitable to optimized. Masterestaurant applies this full checklist in the first week with every group and repeats it each quarter, because channel mix degrades if it is not pushed. Diego F. Parra sums it up in every engagement: what has no 'done' criterion is not done, it is assumed, and assumptions in delivery cost cash every day.

Box 8: fewer than 6 boxes checked with proof is a red alarm

This week's action is one: print the 8 boxes and check only those you can prove with a number. That exercise alone usually uncovers thousands of dollars of hidden margin.

Free tools to apply this now

Masterestaurant tools & method

Diego F. Parra — International consultant, expert in creating and scaling restaurants and in AI applied to restaurants, foodtech and HORECA. Methodology applied in 8.400+ restaurants across 43 countries · Expert in Artificial Intelligence applied to restaurants, hospitality and food businesses · 20+ years in restaurants, catering, large events and business growth · Author of the book «From Slave to Owner» (Amazon) · International keynote speaker for the HORECA sector.

Full profile →

Sector data 2026 (official sources)

Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.

Related content