Ghost Kitchen Menu Planning: Design a Menu That Sells Online

A ghost kitchen menu is not a restaurant menu with the dining room removed. It's an entirely different product — one that has to sell itself through a thumbnail, a one-sentence description, and a price, all on a phone screen at 7 PM to someone who's already hungry and three scrolls deep on DoorDash.

If you're running a delivery-only operation (or thinking about starting one), the menu is the single most important thing you'll build. Get it right and you'll hit the industry benchmark of 20% menu-view-to-order conversion. Get it wrong and you'll watch a 30% platform commission eat whatever margin you had left.

This guide is the playbook for designing a ghost kitchen menu that actually sells. We'll cover the right menu size, which foods actually travel, how to run multiple virtual brands from one kitchen, pricing math that accounts for platform fees, and why your photos are doing 80% of the selling. The options you pick now decide whether your ghost kitchen runs profitable in six months or closes quietly.

Quick Summary: A winning ghost kitchen menu has 15–25 tightly engineered items, each with a professional photo and a specific, sensory description. Professional photography alone can lift delivery app orders 25–70%, but the whole system — menu size, travel-friendly food, margin-aware pricing, and multi-brand strategy — is what separates operators who survive from those who don't.

Why a Ghost Kitchen Menu Isn't Just a Restaurant Menu Online

In a brick-and-mortar restaurant, the menu is one part of a whole experience. The host seats you. The smell of the kitchen drifts past. A server suggests the special, upsells a starter, talks you into dessert. Lighting, music, and plating all contribute to the sale.

None of that exists on Uber Eats.

On a delivery app, your menu is competing against dozens of other ghost kitchens and restaurants inside a cuisine tag, displayed as a scrollable grid of thumbnails. A hungry customer spends around two seconds deciding whether to tap your listing or keep scrolling. Once they're in, they spend another 30–60 seconds scanning items before they bounce or add to cart. Everything you would normally communicate through ambiance, service, and smell now has to happen through two elements: a photo and a sentence.

That's it. Photo and sentence.

This changes how you design the ghost kitchen menu from the ground up:

• No upselling by a server. The menu has to do its own upselling through bundles, combos, and smart item order.
• No signature dish hidden at the bottom. Top-performing items need to be in the first screen of the listing.
• No "I'll explain what it is when you order." Every item has to be self-explanatory in one glance.
• No dine-in portion forgiveness. Food has to arrive 20 minutes later looking and tasting like it just left the pass.

The ghost kitchen operators who treat this as a constraint to optimize — rather than a watered-down version of their real menu — are the ones who make it work.

Keep the Ghost Kitchen Menu Small: 15–25 Items Maximum

The biggest rookie mistake is porting a full dine-in menu with 60 items onto a delivery app. More choice sounds like a good thing. On delivery, it's almost always a mistake.

Behavioral research on menu choice has been clear for decades: too many options increase decision fatigue, slow down ordering, and push customers toward defaulting to whatever they've had before — which, for a new listing, means nothing. Sheena Iyengar's famous "jam study" (24 options vs. 6) showed purchase rates dropping roughly tenfold with more choice. That pattern holds on delivery apps.

Beyond the psychology, a small ghost kitchen menu wins operationally:

• Faster ticket times. Fewer SKUs means cooks build muscle memory. Delivery platform algorithms reward kitchens that prep in under 10 minutes.
• Lower food cost. Fewer ingredients, less spoilage, better per-unit pricing on bulk orders.
• Cleaner photography. 20 items you can shoot beautifully beats 60 items shot inconsistently.
• Higher quality control. Every dish gets the attention it deserves.

The 15–25 item sweet spot looks like this in practice:

• 3–5 hero items (your signature, highest-margin dishes — these do 60% of revenue)
• 4–8 variations of heroes (different proteins, sauces, spice levels on the same base)
• 3–5 sides (fries, salads, rice, bread — high-margin, low-prep)
• 2–4 combos or bundles (these lift average order value)
• 2–3 drinks (canned sodas, bottled water, maybe a signature drink)
• 1–2 desserts (optional, only if they travel well)

Every 60 days, audit performance. Cut the bottom 20% by sales. Iterate the middle. Double down on the top. The menu should be a living document, not a fixed asset.

Menu Engineering for Delivery: What to Put On (and Keep Off) the Menu

Menu engineering is the practice of choosing items based on their operational and financial performance — not just what tastes good. For a ghost kitchen, the formula has three legs: it has to travel, it has to have margin, and it has to survive the packaging.

The delivery-ready filter. Every dish you're considering has to pass the 20-minute test: does it still look and taste like itself after 20 minutes in an insulated bag? If the answer is no, you have two choices — redesign the dish or leave it off the menu.

Foods that travel well

Certain categories are delivery workhorses for a reason:

• Pizza. The blueprint for delivery food. Heat retention, structural integrity, travels flat, reheats tolerably.
• Fried chicken. Wings, tenders, sandwiches. Texture holds if ventilated properly.
• Burgers and smash burgers. Stays hot, packaging is standard, texture forgiving if you use proper buns (brioche beats ciabatta).
• Rice and grain bowls. Poke bowls, Korean bibimbap-style builds, grain bowls. Hold temperature and shape.
• Wraps and burritos. Self-contained, no spill risk, easy to pack.
• Wings. Highest margin per gram of food on most ghost kitchen menus.
• Noodles and pasta. Cook slightly under, oil-based sauces preferred, package sauce separately for delicate cream-based dishes.
• Curries, stews, and soups. Ironically travel beautifully because they're already designed to hold heat in liquid.

Foods that struggle

• Anything breaded that needs crunch. Fries, tempura, schnitzel. Soggy within 15 minutes unless you use vented packaging and steam vents.
• Cream-sauced pastas. Tend to gel and break.
• Leafy salads with dressing applied. Wilt fast. Always package dressing separately.
• Soft-shell tacos. Disintegrate. Use hard shells or rework as build-it-yourself kits.
• Ice cream and anything melty. Don't even try unless you've got dry ice logistics.
• Runny eggs. Continue cooking in transit.
• Crispy-on-bottom-soft-on-top dishes. Loses the textural contrast that made the dish good.

If you love a dish that doesn't survive delivery, resist the urge to shoehorn it onto the menu. Either redesign it (swap sauce, reshape, split components) or save it for catering and pop-ups.

The margin math

Platform commissions of 15–30% mean your food cost needs to land in a different place than a dine-in restaurant. Rough targets for a healthy ghost kitchen:

• Food cost: 25–30% (vs. 30–35% typical for dine-in)
• Packaging cost: 3–5%
• Platform commission: 15–30%
• Labor: 20–25%
• Rent + utilities: 5–10%
• Target margin: 10–15% net

Anything under 65% gross margin on a dish is a red flag. Wings, pizza, rice bowls, and bundled combos tend to hit 70%+. Fine-dining imports tend to hit 40–50% — which is why they don't work in this model.

Packaging is part of the menu

You can't design the food without designing the box it ships in. A perfect dish packaged badly arrives worse than a mediocre dish packaged well. A few non-negotiables:

• Vent holes on anything crispy — traps steam that turns your food into a sponge.
• Separate compartments for wet and dry. Never let sauce pool on fries or bread.
• Structural lids that stack. Drivers stack bags. If your lid caves under another order, you'll eat the refund.
• Leak-proof seals for liquids. Soups, sauces, drinks. Double-seal and label.
• Branded without text on the food. Package can be branded, but DoorDash will reject photos with visible logos on dishes.

Share ingredients across items

One protein should show up in 3–5 items. If you're running a fried chicken concept, your chicken powers tenders, sandwiches, wings, salads, and bowls. Cheese, bread, and sauces pull double or triple duty. This reduces waste, improves prep speed, and protects your cash flow.

Running Multiple Virtual Brands From One Kitchen

Here's where ghost kitchens become interesting. Because there's no physical storefront, one kitchen can power multiple delivery app listings — each with its own brand, menu, and photography. This is how operators scale from one concept to ten without renting new space.

The logic is straightforward. Customers search delivery apps by cuisine: "burgers near me," "healthy bowls," "late-night wings." By running three narrow-focused virtual brands instead of one broad listing, you show up in three cuisine tags instead of one — and each brand speaks directly to a specific craving. Customers find you because each listing is laser-targeted to one search intent, not watered down trying to be everything.

Real-world examples:

• C3 operates Krispy Rice, Umami Burger, and Sam's Crispy Chicken out of shared kitchens — three completely different brands with different photography, prices, and menus, prepared by the same team.
• MrBeast Burger launched by licensing kitchen space in 1,000+ existing restaurants, running a burger brand entirely on top of other operators' capacity.
• The Burger Den is Denny's virtual brand — a burger-focused listing prepared in Denny's kitchens for the late-night delivery crowd.

How to structure a multi-brand ghost kitchen operation:

1. Start with a shared ingredient base. Pick a protein or cuisine family you can execute well, then split it across narrow concepts. Chicken can become: a crispy-chicken-sandwich brand, a wings brand, and a rice-bowl brand. Same prep, three listings.
2. Give each brand distinct visual identity. Same dish photographed on a dark moody background for brand A and a clean white background for brand B will read as two entirely different restaurants. This is non-negotiable — customers cross-reference listings, and identical photography destroys credibility. Our AI food photography for ghost kitchens guide covers how to build distinct brand libraries fast.
3. Write brand-specific copy. The wings brand is playful and loud. The rice bowl brand is clean and healthy. The sandwich brand is confident and simple. Voice differentiates.
4. Keep menus narrow per brand. 10–15 items each. Each brand exists to dominate one cuisine tag, not be everything to everyone.
5. Avoid the transparency trap. Customers are increasingly savvy — many know ghost kitchens run multiple brands. Don't hide it, but don't make it obvious either. Different names, different photography, different voice is enough.

When should you launch a second brand? When your first one is stable, your prep capacity is under 70%, and you can identify a demand gap in your local delivery market that your existing kitchen can fill without slowing service.

Your Photos Are Your Storefront (Treat Them That Way)

On a delivery app, your photo is not decoration. It's the entire storefront.

There's no window display, no A-frame sign, no smell of fresh bread pulling people in. The only visual cue a customer has before ordering is the image that represents your food. And the data on what photos do is overwhelming:

• Grubhub: Adding food photos increases orders by up to 30%. Restaurants with photos and descriptions see up to 70% more orders than text-only menus.
• DoorDash: Items with photos generate up to 44% more monthly sales. Header images lift sales by up to 50%.
• Deliveroo: Professional photos drive a 24% order increase.
• Just Eat: Quality photos deliver 4x more basket additions.
• Industry average: Menu conversion rate rises by 25% or more when menus use images versus text-only.

And the gap is widening. DoorDash's 2024 trends report showed an 11% year-over-year increase in consumer reliance on food photos when making ordering decisions. For Gen Z — the fastest-growing delivery demographic — 46% say menu photos directly influence what they order.

The thumbnail test. Before you upload any photo to a delivery app, shrink it to 60 × 60 pixels and look at it. Can you tell what the food is? Is there one clear focal point? Is the color contrast strong enough to catch a scrolling eye? If not, reshoot. Your photos will live as thumbnails far more often than they'll live at full size.

Every item needs a photo. Not just top sellers. Deliveroo's internal data shows that vendors with a larger portion of their menu photographed get more total orders — because unphotographed items look like afterthoughts, and customers subconsciously downgrade the entire listing. If you're launching with 20 items, plan to photograph all 20.

Platform specs matter.

• DoorDash: 16:9 landscape, minimum 1400 × 800 pixels, strict 14-category rejection policy
• Uber Eats: 5:4 to 6:4 landscape, minimum 1200 × 800 pixels
• Grubhub: Landscape HD, DSLR quality recommended

Using the wrong aspect ratio is the #1 reason menu photos get rejected. Shoot with breathing room around the plate so you can export two different crops from one master file. For a full platform-by-platform breakdown, see our guide on food photography for delivery apps and the detailed Uber Eats photo requirements.

Brand consistency across a virtual brand. Every dish in one virtual brand should feel like part of the same visual universe. Same plate style, same background, same lighting temperature, same angle convention (overhead for flat dishes, 45° for stacked). Mix styles and the brand looks like a franchise with no creative direction.

The realistic path for most operators. Traditional food photography costs $100–$200 per dish plus a full shoot day. For a 20-item menu across three virtual brands (60 total photos), that's $6,000–$12,000 per round — and you need to re-shoot whenever you add or tweak items. AI-powered photography tools have compressed that to minutes per dish, which is why most modern ghost kitchen operators treat photography as an ongoing production line, not an annual event. If you want a deeper dive into the workflow, the restaurant food photography guide walks through the practical side.

Writing Descriptions That Sell When Customers Can't See or Smell the Food

Your photo stops the scroll. Your description closes the sale.

A menu description on a delivery app has about 150–250 characters of prime real estate before the "read more" link. That's roughly two sentences. Waste one and you've cut your pitch in half.

The strongest ghost kitchen menu descriptions follow a simple three-part structure:

1. Hook: lead with the flavor or experience promise. Crispy Nashville hot chicken, slick with honey and cayenne.

2. Build: specific ingredients and prep. Brined 12 hours, double-dredged, fried to order, tossed in our in-house hot sauce.

3. Close: portion, pairing, or texture cue. Served on buttered brioche with bread-and-butter pickles. One sandwich; enough to make you quiet for a minute.

All three in roughly 50 words. Specific. Sensory. Ends with a reason to tap.

What weak descriptions sound like:

• "Delicious chicken sandwich made with high-quality ingredients."
• "Our signature burger, loved by customers."
• "Fresh vegetables and homemade sauce."

These say nothing. They're the verbal equivalent of a blurry photo. Every delivery app is full of them, which is exactly why yours should be different.

Rules that consistently work:

• Specific verbs beat adjectives. "Braised 12 hours" is stronger than "slow-cooked." "Pressed under cast iron" beats "grilled."
• Name the ingredients that matter. "Manchego and local honey" beats "premium cheese." People search for specific ingredients to find what they want.
• Include texture and temperature. "Crispy," "molten," "ice-cold," "charred." The senses that can't be conveyed otherwise.
• Mention portion context without being awkward. "Two-handed sandwich." "Sharable for two." "One handheld." Helps customers calibrate the order.
• Slip in searchable terms naturally. Delivery app search is keyword-matched. If your dish is vegan, gluten-free, spicy, or Korean — say those words. Don't rely on tags alone.
• Avoid cliché copy. "A symphony of flavors." "Mouthwatering." "The best in the city." These have been used so often they've become invisible.

A rewrite example:

Before: "Our famous Chicken Tikka Masala — a rich and flavorful Indian curry with tender chicken pieces."

After: "Tandoor-charred chicken thighs simmered in a tomato-cream sauce with toasted cumin and kashmiri chili. Rich, smoky, medium-heat. Comes with a mound of basmati and warm naan."

The second version gives you a clear taste preview, a heat level, and a portion picture — in the same character count.

Pricing Strategy: Protecting Margin After Platform Fees

Pricing a ghost kitchen menu is where most operators lose money quietly. The margin math on delivery looks brutal once you factor in every real cost, and the only way to survive is to price for the full economics, not just food cost.

The honest breakdown on a $30 order:

That's the best case. Miss on food cost, run heavy promos, or lean on the Premier commission tier and you're at break-even or below.

What to do about it:

• Price 10–20% above dine-in equivalents. This is now standard practice on delivery apps. Customers expect a delivery premium.
• Build bundles that lift AOV above $25. Add-on logic on DoorDash and Uber Eats is aggressive — customers who start a cart will add sides, drinks, and desserts if the economics are compelling. Combos do this automatically.
• Use high-margin heroes to subsidize traffic drivers. Wings, rice bowls, and sides often hit 75%+ margin. Pizza and sandwiches are often 65–70%. A mix of both protects the blended margin.
• Avoid deep platform promos on low-margin items. A 20% off discount on a 65% margin dish is survivable. The same discount on a 50% margin dish is bleeding.
• Raise prices 3–5% twice a year. Customers rarely notice incremental increases. They absolutely notice when you raise everything at once by 15%. Steady small increases keep you ahead of cost inflation.

Bundle architecture. The goal is to anchor the cart above the $20 psychological threshold where customers start adding sides and drinks. A combo that pairs a hero item with a side and a drink at a $3 discount off the individual total will usually outperform the same items sold separately — because you've done the math for the customer.

Optimizing the Menu With Delivery Platform Analytics

Once the ghost kitchen menu is live, the real work begins. Ghost kitchen menus are not set-and-forget — they're living systems that should be tuned every month based on hard data.

Metrics that matter:

• Menu view → order conversion rate. The single most diagnostic number. A healthy rate is around 20% across delivery apps. Below 12% means your photos, prices, or descriptions are failing at first contact.
• Item-level conversion rate. Which dishes get viewed but not added? Those are your photo/description optimization targets.
• Average order value (AOV). Trend line more important than absolute number. Rising AOV usually means combos and upsells are working.
• Repeat order rate. Customers who ordered from you more than once in the last 60 days. Healthy ghost kitchens run 25%+.
• Star rating and review sentiment. Below 4.5 stars and your listing falls in search rank. Packaging complaints and temperature complaints are the top killers.
• Cancellation rate. Over 3% is a red flag — usually a prep-time or accuracy problem.

Monthly audit cadence:

1. Cut the bottom 20% by sales. They take up photography budget and mental overhead without earning their place.
2. A/B test the top 5. Swap one element at a time — a new photo, a rewritten description, a price nudge. Run it for 2 weeks, compare.
3. Expand what's winning. If a burger is crushing, add two variations. If a bowl is lagging, don't launch three more — fix the one you have.
4. Read customer photos. Customers upload photos of what they received. If yours look different from your menu photos, you have a plating consistency problem. Fix it before it becomes a review problem.
5. Watch seasonality. Winter menus lean heavier, spicier, richer. Summer skews lighter and fresher. Plan two menu refreshes per year — a spring/summer version and a fall/winter version. If you want tactical ideas for driving traffic beyond the menu itself, see our guide on cloud kitchen marketing strategies.

Platform tools to use:

• DoorDash Merchant Portal — item-level performance, promo analytics, customer feedback
• Uber Eats Manager — similar insights with conversion funnel visibility
• Grubhub for Restaurants — item performance and rating trends

Pull the data weekly. Make changes monthly. Let six months of iteration compound.

A 30-Day Action Plan to Rebuild Your Ghost Kitchen Menu

If you're starting from a menu that's been on autopilot, here's how to rebuild it in four weeks:

Week 1 — Audit. Pull last 90 days of data from every delivery platform. Sort items by sales and conversion rate. Make three lists: keep (top 40%), fix (middle 40%), cut (bottom 20%). Check margin on every "keep" item. Anything under 65% gross margin gets reworked or repriced.

Week 2 — Rewrite. Every description gets the three-part treatment (hook → build → close). Target under 200 characters. Read each one out loud. If it sounds generic, rewrite it. For a framework on getting discovery as well, our guide to getting more orders on DoorDash pairs well with menu-level fixes.

Week 3 — Reshoot. Photograph every single item on the "keep" and "fix" lists. Use consistent plating, background, and lighting across an entire virtual brand. Export in 16:9 for DoorDash and 5:4 for Uber Eats. If you're resource-constrained, this is exactly the point where AI food photography tools earn their keep — you can go from a phone snap to a professional-grade menu image in about 90 seconds per dish. Our menu photoshoot guide covers the full planning workflow either way.

Week 4 — Relaunch and baseline. Upload the new ghost kitchen menu across every platform. Track conversion rate, AOV, and star rating daily for the first two weeks. Establish your new baseline. Don't make any more changes for 30 days — give the algorithm time to redistribute visibility.

Ongoing — 60-day review cadence. Every two months, repeat the audit. Cut, fix, keep. Test one new item against one existing item. Let the data make the decision.

Frequently Asked Questions

How many items should a ghost kitchen menu have?

15–25 items per virtual brand is the sweet spot. Enough variety to cover different cravings and price points, tight enough to maintain operational speed and photo quality. Ghost kitchens running 40+ item menus almost always see slower ticket times, higher food waste, and weaker overall ratings. If you feel the urge to add a 26th item, cut the lowest-performing existing item first.

What foods travel best for delivery?

Pizza, fried chicken, burgers, rice and grain bowls, wraps, burritos, wings, pasta (with oil-based sauces or sauce-on-the-side for cream), curries, and stews. Avoid anything that relies on textural contrast between crispy and soft elements, cream-sauced pastas, leafy salads with dressing already applied, soft-shell tacos, runny eggs, and anything frozen or melted.

How much should I raise prices for delivery apps?

Most ghost kitchens and restaurants with delivery operations price 10–20% above their dine-in equivalents on delivery apps. This offsets the 15–30% platform commission while keeping prices competitive against nearby listings. Check the 5–10 restaurants in your cuisine tag and position at or slightly below their pricing — unless you have a clear differentiator (better photos, faster delivery, tighter menu).

Do I really need a photo for every menu item?

Yes. Unphotographed items get ordered 3–5x less than photographed ones, and having unphotographed items visible on the listing subconsciously downgrades the entire brand in customers' eyes. If budget is tight, prioritize your top 10 sellers first, but plan to have every item photographed within 30 days of launch.

Can one ghost kitchen run multiple virtual brands?

Yes, and it's one of the main advantages of the model. Large operators like C3 run three or more brands (Krispy Rice, Umami Burger, Sam's Crispy Chicken) from shared kitchens. Independent operators commonly run two or three brands — often a hero concept plus a sub-brand that monetizes spare capacity (e.g., a burger brand that also runs a wings brand for late-night orders). Each brand needs distinct photography, voice, and a narrow menu focus.

How often should I update my ghost kitchen menu?

Minor optimization continuously. Major refreshes twice a year — a spring/summer and a fall/winter version. A full audit every 60 days where you cut the bottom 20% of sellers and expand what's working. Delivery apps reward actively managed listings with higher visibility in search, making it easier for customers to find your ghost kitchen.

What's a good conversion rate on DoorDash or Uber Eats?

Roughly 20% menu-view-to-order conversion is a healthy benchmark across delivery platforms. Top-performing listings hit 25–30%. Anything under 12% signals a problem with photos, prices, or first-screen descriptions. Platform dashboards (DoorDash Merchant Portal, Uber Eats Manager) show this by item and at the listing level.

Should I write my own menu descriptions or hire a copywriter?

Write them yourself first — you know the food better than anyone. Use the three-part structure (hook → build → close), keep each one under 200 characters, and read them out loud. If you're running five or more virtual brands or don't enjoy writing, a food copywriter can speed up the process. Budget $15–$40 per description for a specialist. Either way, rewrite every description at least once per year to keep them fresh.

---

Ghost kitchen menus reward discipline. Tight item count, food that actually travels, photography that earns every tap, descriptions that close the sale, prices that survive platform fees, and a monthly habit of cutting what isn't working.

None of this is glamorous. All of it compounds. The ghost kitchen operators who treat their menu like a product they ship and iterate — rather than a fixed document they printed once — are the ones running profitable kitchens in 2026.