A delivery menu should not be a digital copy of the dine-in menu.
The guest experience is different. The food spends time in a container. The restaurant pays packaging cost. The order may carry platform fees, promotions or refund risk. Some items travel well; others arrive soft, separated, cold or damaged. High-volume delivery can also overload a kitchen station that works perfectly during dine-in service.
Delivery menu engineering asks a different question from standard menu engineering:
Which items create strong contribution after delivery-specific costs while still arriving in a condition the restaurant is willing to put its name on?
The goal is not to offer everything. The goal is to build a delivery assortment that is operationally reliable and economically deliberate.
Start with a separate delivery item list
Export the items currently offered through delivery channels.
For each item, record:
• selling price,
• ingredient cost,
• packaging cost,
• platform or channel cost where measurable,
• discount or promotion exposure,
• average preparation time,
• refund or complaint history,
• travel-quality risk,
• sales volume.
Do not begin with menu design language such as stars or dogs. First build the delivery economics.
Calculate item contribution before delivery-specific costs
The basic contribution calculation is:
Item contribution = Selling price – direct item cost
If a dish sells for $18 and its ingredient cost is $5.40:
$18 – $5.40 = $12.60
That number is not yet the delivery contribution.
Delivery creates additional direct costs that may not exist in dine-in.
Add packaging to the item cost
Packaging should be assigned at the order or item level in a consistent way.
Possible costs include:
• entrée container,
• side container,
• sauce cup,
• lid,
• bag,
• tamper seal,
• cutlery,
• napkins,
• drink carrier.
Example:
• ingredient cost: $5.40,
• item packaging: $1.10,
• allocated bag/seal cost: $0.35.
Direct product-and-packaging cost:
$5.40 + $1.10 + $0.35 = $6.85
If price remains $18:
Pre-channel contribution = $18 – $6.85 = $11.15
The figures are illustrative.
Treat platform economics separately from food cost
Third-party platform fees, commissions, advertising and promotions should not be hidden inside ingredient food cost.
For management analysis, show them separately.
A useful structure is:
Delivery contribution before fixed overhead = Selling price – ingredient cost – packaging – channel-variable costs
The exact commercial treatment depends on the restaurant’s contracts and accounting model.
Do not publish or rely on a generic platform-fee percentage. Use the restaurant’s actual statements.
Compare delivery and dine-in contribution
An item can have the same recipe cost in both channels but different economics.
Example:
| Measure | Dine-in | Delivery |
|---|---|---|
| Selling price | $18.00 | $19.50 |
| Ingredient cost | $5.40 | $5.40 |
| Packaging | $0.20 | $1.45 |
| Channel-variable cost | $0.00 | $3.10 |
| Contribution before fixed overhead | $12.40 | $9.55 |
The figures are illustrative.
The right response is not automatically to raise delivery price. The restaurant should evaluate guest demand, platform rules, competitive position and the total order.
Evaluate travel quality
Economics do not save an item that consistently arrives badly.
Test each major item after a realistic travel window.
Review:
• temperature,
• crispness,
• moisture migration,
• sauce separation,
• bread texture,
• garnish condition,
• portion movement,
• packaging leakage,
• presentation.
Test the item the way the guest receives it—not immediately after plating.
Create a travel-quality score
A simple internal score can help.
Example:
| Factor | Score 1-5 |
|---|---|
| Temperature retention | 4 |
| Texture | 2 |
| Packaging stability | 5 |
| Appearance | 3 |
| Reheat tolerance if relevant | 4 |
Do not turn the score into fake precision.
Its purpose is comparison.
An item that is highly profitable but scores poorly on travel quality may need redesign, channel restriction or removal.
Engineer recipes for delivery when the brand allows it
The delivery version does not always need to be identical to dine-in if the restaurant can preserve the product promise.
Possible changes include:
• sauce packed separately,
• garnish added differently,
• crisp components separated,
• bread toasted to a different level,
• fries replaced with a more travel-stable side,
• portion adjusted to fit packaging,
• cold and hot components separated.
Any variation should be standardized and costed.
Do not let staff invent a new delivery build during service.
Protect the packaging-product relationship
Packaging is part of the delivery recipe.
A container that is too small can damage presentation. A sealed hot item can trap steam. A vented container may help texture but reduce heat retention. A tall item may move during transport.
Test the food and package together.
Keep a controlled packaging specification by item group.
Review item preparation time
Delivery volume can change kitchen economics through labor and throughput.
An item may have strong contribution but require:
• long cook time,
• multiple stations,
• hand finishing,
• scarce equipment,
• complex modifiers.
During peak demand, that item can slow other orders.
Measure actual ticket-time effect.
A delivery menu should protect kitchen capacity, not simply maximize the number of items listed.
Identify station bottlenecks
Map each delivery item to its primary production station.
Example:
| Item group | Main station | Peak concern |
|---|---|---|
| Burgers | Grill | Grill capacity |
| Fried chicken | Fry | Fryer queue |
| Pizza | Oven | Oven slots |
| Salads | Cold line | Assembly labor |
| Desserts | Pantry | Packaging |
If one channel drives an excessive share of demand to one station, consider:
• limiting the delivery assortment,
• adjusting item availability,
• moving prep earlier,
• changing menu design,
• controlling promised order times.
Separate high-volume from high-value items
An item that sells often is not automatically valuable.
Review at least:
• number sold,
• contribution dollars per item,
• total contribution dollars,
• complaint/refund rate,
• operational load.
An item with moderate contribution but huge volume may still be important.
Another item may have high unit contribution but create little total value because it sells rarely.
For standard item-level profitability: Restaurant Menu Item Profitability: How to Measure Contribution Margin
Add refund and remake cost to the review
Delivery problems can create direct loss.
Track by item:
• refund frequency,
• remake frequency,
• missing-component complaints,
• quality complaints,
• packaging failures.
Do not assume every refund means the recipe is bad. The cause may be:
• order accuracy,
• driver delay,
• packaging,
• platform issue,
• kitchen execution.
But repeated item-level problems belong in menu engineering.
Review modifier complexity
Delivery menus often accumulate modifiers because the digital platform makes options easy to add.
Too many modifiers can create:
• slower ordering,
• kitchen errors,
• longer prep time,
• packaging complexity,
• inconsistent pricing,
• missing components.
Review whether each modifier:
• changes guest value,
• has a clear price,
• has a controlled food cost,
• can be executed accurately,
• belongs on every channel.
A delivery menu should not become a database of every possible request.
Price paid modifiers deliberately
If an extra protein portion costs $2.20 and sells for $3.00, the incremental contribution is $0.80 before delivery-specific costs.
That may be weak compared with the operational burden.
Use contribution, not markup intuition.
For a broader review of add-ons and extras, use the restaurant’s modifier-profitability process when available.
Remove items that create channel-specific failure
An item can be excellent in the dining room and wrong for delivery.
Possible removal triggers:
• persistent travel-quality failure,
• high refund rate,
• packaging failure,
• weak contribution after channel costs,
• severe station bottleneck,
• excessive preparation time,
• low demand and high complexity.
Removal is not an admission that the dish is bad.
It means the channel is different.
Use delivery-only bundles carefully
Bundles can improve order value and simplify decisions.
But the bundle should be tested for:
• total contribution,
• packaging,
• modifier behavior,
• production load,
• promotion cost,
• item substitution.
Do not judge the bundle only by gross sales.
If a value offer increases revenue but shifts demand toward low-contribution components, it can weaken the order economics.
Build price architecture by channel
Some restaurants use the same price across channels. Others use different pricing where contracts and platform rules allow.
Whatever the decision, keep it deliberate.
Document:
• dine-in price,
• direct online price,
• third-party delivery price,
• promotional price,
• effective net revenue.
The restaurant should understand the reason for any difference.
Protect price consistency across systems
A delivery menu becomes difficult to manage when prices or availability differ unintentionally between:
• POS,
• restaurant website,
• delivery platforms,
• menu boards,
• printed materials.
Online-ordering integration reduces that risk: Online Ordering Integration: Connect Orders Without Creating Operational Chaos
The menu-engineering process should define what the correct price and availability should be. Technology should then distribute it reliably.
Analyze the whole order, not only each item
Delivery economics also depend on basket composition.
Useful measures include:
• average order value,
• items per order,
• beverage attachment,
• side attachment,
• dessert attachment,
• packaging cost per order,
• discount value per order,
• refund value per order.
A low-contribution entrée may still support a strong basket if it reliably drives profitable add-ons.
Use order-level data before removing a high-volume item.
Test minimum order economics where applicable
If the restaurant controls its own ordering channel, minimum-order rules can protect delivery economics.
But any threshold should be tested against guest behavior and local expectations.
The objective is to avoid orders where packaging, payment, delivery or handling costs overwhelm contribution.
Review promotions by net contribution
A delivery promotion can create impressive order counts and weak economics.
For every promotion, compare:
• incremental orders,
• selling price,
• discount value,
• ingredient cost,
• packaging,
• platform or ad cost,
• incremental labor/capacity effect.
For a broader promotion framework: Restaurant Promotion Profitability: Measure Discounts, Traffic and Contribution Before Repeating an Offer
Build daypart availability if demand differs
The best delivery menu at lunch may not be the best menu late at night.
Use daypart controls where the operation supports them.
Examples:
• remove slow breakfast items after a defined time,
• simplify the late-night menu,
• restrict complex items during peak dine-in periods,
• enable catering-style family meals only in appropriate windows.
Availability should follow kitchen capacity and guest demand.
Use sold-out controls aggressively
Selling an unavailable item creates:
• guest disappointment,
• refund risk,
• staff calls,
• substitutions,
• platform friction.
The operation needs a fast process for:
• marking sold-out items,
• syncing status,
• identifying substitutes,
• restoring availability.
Delivery menu accuracy is part of the product.
Run controlled travel tests
For a major item:
- Prepare it exactly as during service.
- Pack it in the approved packaging.
- Hold it for the typical travel interval.
- Move it as a real order would move.
- Open it at the end of the test interval.
- Assess temperature, texture, appearance and leakage.
- Record the result.
- Test one controlled change if needed.
Repeat for high-volume or complaint-prone items.
Build a delivery item scorecard
A practical scorecard might contain:
| Measure | Why it matters |
|---|---|
| Selling price | Revenue |
| Ingredient cost | Product economics |
| Packaging cost | Channel cost |
| Variable channel cost | Net economics |
| Unit contribution | Profitability signal |
| Volume | Demand |
| Travel score | Guest quality |
| Prep time | Capacity |
| Refund rate | Failure cost |
| Station | Bottleneck visibility |
Use it to rank decisions, not to create a decorative dashboard.
Use four decision outcomes
Each item can end in one of four actions.
Keep
Strong economics and reliable delivery performance.
Fix
Demand is useful, but packaging, recipe, price or execution needs improvement.
Restrict
Works only in certain dayparts, distances or channels.
Remove
Weak economics and weak guest outcome with no practical fix.
This keeps menu engineering operational.
Delivery-menu review workflow
- Export the current delivery assortment with price, sales and modifier structure.
- Add ingredient, packaging and channel-variable costs.
- Calculate contribution using consistent cost definitions.
- Add travel-quality results for high-volume and high-risk items.
- Map each item to its main production station and identify bottlenecks.
- Review complaints, refunds and remakes for repeated item-level failure.
- Compare basket behavior so items are not judged in isolation.
- Assign a clear outcome: keep, fix, restrict or remove.
- Update price and availability across every ordering channel.
- Recheck sales, contribution and quality data after four to six weeks.
Run a first delivery-menu audit
- Select the 10 highest-volume delivery items and pull price, units sold, refunds and modifier data for one completed period.
- Add current ingredient cost, item-level packaging and actual variable channel costs using the restaurant’s own statements.
- Calculate contribution before fixed overhead for each item using one consistent definition.
- Run a realistic travel test for high-volume or complaint-prone items and record temperature, texture, appearance and packaging failures.
- Map each item to its main station and identify whether delivery volume creates a peak-period bottleneck.
- Classify each item as keep, fix, restrict or remove based on economics, travel quality and operational load.
- Update price, modifiers and availability across every ordering system at the same time.
- After four to six weeks, compare contribution, refund rate, basket behavior and ticket impact with the original baseline.
Common delivery-menu mistakes
Copying the dine-in menu
The channel has different economics and quality constraints.
Ignoring packaging cost
Item contribution is overstated.
Ignoring platform statements
Gross sales look stronger than net economics.
Too many modifiers
Order accuracy and speed suffer.
Keeping low-demand complexity
The menu becomes harder to execute.
No travel test
Management evaluates food at the pass instead of at the guest’s door.
No station-capacity view
A profitable item can still damage throughput.
Frequently asked questions
Should every dine-in item be available for delivery?
No. Offer items that maintain acceptable quality, economics and operational reliability in the delivery channel.
Should delivery prices be higher?
That depends on contracts, platform rules, market conditions and the restaurant’s economics. The important control is understanding net contribution by channel.
How often should the delivery menu be reviewed?
Review after meaningful price, packaging, platform, menu or demand changes, and on a regular cadence appropriate to delivery volume.
What is the first metric to calculate?
Start with contribution after ingredient and packaging cost, then add actual channel-variable costs and quality data.
Is travel quality more important than margin?
Both matter. A high-margin item that repeatedly disappoints guests can create refunds and damage the brand. A perfect traveler that loses money is also not sustainable.
Next action
Choose the ten highest-volume delivery items. Add ingredient cost, packaging cost, actual channel-variable cost, travel-quality result and refund history. If an item looks strong only before delivery costs or only before travel testing, it is not yet a strong delivery item.