Key takeaways
- Off-premises now matters more: 58% of limited-service and 41% of full-service operators say it is a bigger share of sales than in 2019.
- Score each dish on travel, speed, margin and packing before it goes on the delivery menu.
- A shorter delivery menu with bundles usually sells more and causes fewer refunds than a full copy of the dine-in menu.
- Over 80% of consumers use deals such as combos, so build family meals and bundles on purpose.
Start from a different question
A dine-in menu answers the question "what do we cook best?" A delivery menu has to answer "what do we cook best that still tastes right twenty minutes later in a box, can be packed fast and earns its margin after fees?" Many restaurants skip that second question and copy the full dine-in menu onto their ordering page and delivery apps. The result is predictable: soggy dishes, long tickets at peak, packing mistakes and refunds.
Getting it right matters more each year. 58% of limited-service and 41% of full-service operators say off-premises is a larger share of their sales than in 2019[1]. Among full-service operators that offer delivery, 57% say it is a bigger share of sales than before the pandemic[2].
Score every dish before it goes on
Use a simple scoring rubric so decisions are based on evidence rather than favorites. Score each dish from 0 to 2 on five criteria, with a travel test behind the first score.
| Criterion | What a 2 looks like | What a 0 looks like |
|---|---|---|
| Travel quality | Tastes and looks right after your longest delivery time | Soggy, separated, cold or crushed on arrival |
| Speed at peak | Built from stations that are not your bottleneck | Needs the busiest station or a long cook at peak |
| Packing simplicity | One container, few components, hard to get wrong | Several containers, many modifiers, easy to miss a part |
| Margin after fees | Earns well even after commission and packaging | Loses money unless priced far above competitors |
| Demand | Sells well in store or is a known off-premises favorite | Rarely ordered, or only by dine-in regulars |
Run the travel test honestly: pack the dish exactly as a guest would get it, leave it in a car for your longest typical delivery time and taste it. Our guide to takeout packaging that travels well has the full test.
Example: a fictional bistro scores six dishes. Its braised short rib scores 9 out of 10: it travels well, holds on the line and earns a good margin. Its crispy calamari scores 4, because it goes soft in transit and needs the fryer at peak. Its chicken Caesar scores 8 once the dressing goes on the side. Its soufflé scores 2 and stays dine-in only. Its burger scores 7 and its seasonal risotto 5, which turns soft and stiff after twenty minutes. The bistro launches with the short rib, Caesar and burger, reworks the calamari as a vented, sauce-on-the-side version to retest, and keeps the risotto and soufflé for the dining room.
Decide what to cut, fix or keep
Use the scores to sort dishes into three groups.
Keep dishes scoring 8 to 10
Put them on the delivery menu as they are, and feature the best sellers first.
Fix dishes scoring 5 to 7
Change the container, send components separately, adjust the recipe for travel or simplify modifiers, then retest.
Cut dishes scoring 4 or below
Keep them for dine-in. A dish that fails in transit costs you a refund and a guest, not just one sale.
Retest after changes
Any recipe, supplier or container change means a new travel test.
Review the whole menu twice a year
Score new items before launch and rescore old ones when complaints or refunds rise.
Cutting feels risky, but a shorter delivery menu is usually easier to choose from, faster to cook and easier to pack correctly. Guests ordering on a phone scroll quickly; twenty strong items do better than sixty mixed ones.
Build bundles and family meals
Over 80% of consumers use deals such as buy-one-get-one offers, combo meals or real-time specials[3]. Bundles turn that habit into a menu feature you control.
- Meal for one: a main, a side and a drink at a small saving over buying them separately.
- Family meal: a large main, two sides and a dessert for four, packed in fewer containers than the same items ordered separately.
- Lunch bundle: a fixed combo for weekday office orders, built from your fastest stations.
- Party pack or catering tray: shareable portions for groups, with a minimum lead time.
Bundles help the kitchen as well as the guest. They concentrate orders on dishes you have already scored well, simplify packing and raise the average ticket. Price them so the saving is visible but the margin holds after commission; see how to price menu items on delivery apps.
Keep modifiers simple
Every modifier is a chance for a mistake. On a delivery menu, keep them to what guests really need.
- Make essential choices required: temperature for burgers, spice level, choice of side.
- Limit optional add-ons to the few that guests actually order.
- Avoid free-text special instructions where possible, or cap their length and make them prominent on the ticket.
- Use the same modifier names on every channel so the kitchen reads one language.
- For allergens, provide clear tags and a note to call the restaurant for serious allergies. The eight major allergens accounted for 90% of US food allergies when federal labeling law passed[4], so make those easy to spot.
Fewer modifiers also mean fewer missing items. Our guide to preventing wrong and missing delivery orders covers packing checks for the modifiers you keep.
Name and order items for a phone screen
Delivery menus are read on a small screen, often with no photo. Names should say what the dish is: "Grilled chicken bowl with rice and greens" beats a clever name nobody understands. Put the categories in the order guests decide: popular items and bundles first, then mains, sides, drinks and desserts. Keep descriptions short and concrete, mentioning portion, key ingredients and anything that might surprise, such as heat or a shared size. For photo and description rules on marketplaces, see delivery app listings that sell.
Seasonal and limited-time items
Limited-time items are a good way to keep a delivery menu fresh without making it longer. Swap one or two items each season rather than adding new ones on top. Score each new item with the same rubric before launch, and give it a clear end date in the listing so guests do not complain when it disappears.
Use limited-time items to test ideas cheaply. A dish that sells well on delivery for a season, with low refunds, has earned a permanent place. One that sells only when promoted, or that draws complaints about temperature or texture, can quietly come off at the end of the run.
Keep the kitchen in mind here too. A seasonal item that needs a new ingredient or a new container adds prep and storage. Where possible, build limited-time items from ingredients you already stock and pack them in containers you already use, so the novelty is on the plate and not in the back of house.
Tell staff about every change before it goes live, including the packing kit and any new modifiers, so the first orders are packed as carefully as the last.
Design around your kitchen's capacity
The best delivery menu draws on stations that have spare capacity when off-premises orders peak. If your fryer is overloaded on Friday nights, a delivery menu heavy on fried items will slow dine-in too. Look at your ticket times by station during off-premises peaks and favor dishes from stations with headroom. Some restaurants add a few delivery-only items built from prepped components, such as bowls or wraps, that can be assembled quickly at a separate station.
Speed is what guests notice first. 94% of consumers say speed is critical when ordering off-premises[5]. A menu that keeps the line moving is a service decision as much as a culinary one.
Price for the channel and check the results
Most operators already charge more on third-party channels: 83% of limited-service and 66% of full-service restaurants do[6]. Your delivery menu should be priced for the channel it is on: menu prices on your own ordering page, and a markup on apps that covers commission, packaging and refunds. Recheck margin for every item after fees, and drop items that only sell when discounted.
Each month, look at four numbers by item: sales, refunds and complaints, average ticket time and margin after fees. Items that rank low on all four are candidates to cut. Items with high sales but rising complaints need a packaging or recipe fix before they cost you guests. Remember that packaging rules vary by place; our guide to foam bans and single-use plastic rules explains what to check before you choose containers for new dishes.
Frequently asked questions
How many items should a delivery menu have?
There is no fixed number, but it should usually be shorter than your dine-in menu. Keep dishes that travel well, cook quickly at peak and earn their margin after fees, and add a few bundles. A focused menu is easier to choose from and to pack correctly.
What foods do not travel well for delivery?
Crisp fried foods in sealed containers, delicate sauces, dishes that need to be eaten immediately such as soufflés, and items that separate or set as they cool, such as some risottos. Test each dish in a real delivery time before deciding.
Should delivery menus have different items from dine-in?
Often, yes. Remove dishes that fail in transit, add bundles and family meals, and consider a few delivery-only items built from prepped components at a station with spare capacity.
How do I test whether a dish works for delivery?
Pack it exactly as a guest would receive it, leave it for your longest typical delivery time, then check temperature, texture and appearance and taste it. Score it on travel, speed, packing, margin and demand.
References
National Restaurant Association (2025). From Trend to Transformation: Off-Premises Dining Now Essential for Restaurant Consumers, Operators (opens in a new tab) ↑
National Restaurant Association (2026). National Restaurant Association comments to the FTC on the Food Delivery Fees ANPRM (Project No. P267101) (opens in a new tab) ↑
National Restaurant Association (2025). From Trend to Transformation: Off-Premises Dining Now Essential for Restaurant Consumers, Operators (opens in a new tab) ↑
US Food and Drug Administration (2004). Food Allergies (opens in a new tab) ↑
National Restaurant Association (2025). From Trend to Transformation: Off-Premises Dining Now Essential for Restaurant Consumers, Operators (opens in a new tab) ↑
National Restaurant Association (2026). National Restaurant Association comments to the FTC on the Food Delivery Fees ANPRM (Project No. P267101) (opens in a new tab) ↑