Three Delivery Models to Choose Between

"Restaurant delivery service" is not one business. Before you write a word of the plan, decide which of three models you are actually building, because each carries a different cost base, a different margin ceiling, and a different funding story. The template forces this decision up front; most weak plans blur it and end up describing a business that cannot make money.

Many operators start on marketplaces to prove demand, then migrate customers to direct ordering to reclaim margin. The single number that decides whether the business works is the direct-versus-app order mix, and your plan should model it explicitly rather than assume a fixed platform fee forever.

Mistakes That Sink Delivery Startups

Delivery is unforgiving because a 25% commission and cold food both attack the same thin margin. These are the errors we see most often when founders bring us a delivery plan to fix:

• Building the whole model on third-party apps. A blended 24% commission on 90% of orders can turn a healthy-looking gross margin into a loss. The plan needs a route off app dependence, not just a launch on it.
• Pricing delivery items the same as dine-in. If a $12 burger costs $12 on the app, the commission is coming straight out of your pocket. Operators typically add a 10-20% delivery-menu markup to offset the take rate.
• No insurance for drivers. Personal car policies exclude commercial delivery. Running uninsured drivers is the fastest way to a claim that ends the business; hired and non-owned auto cover (US) or courier cover (UK) is non-negotiable.
• A menu that does not travel. Crispy items, delicate plating, and anything that steams in a sealed box arrives disappointing. High-review delivery brands engineer the menu around a 20-30 minute transit.
• Cooking from a home kitchen. Most US states and the UK bar delivery of higher-risk cooked food from a domestic kitchen. Assuming you can start from home is the single most common licensing mistake.
• No repeat-order engine. If every order is a fresh paid acquisition through an app, customer acquisition cost never falls. Loyalty, direct offers, and packaging inserts are what compound.

What It Costs to Launch

A delivery-only operation working out of an existing commissary can open for as little as $15,000-$40,000 (£12,000-£32,000). A dedicated ghost kitchen with its own delivery fleet is a different animal at $120,000-$250,000+ (£95,000-£190,000+). The gap between those two numbers is almost entirely about whether you rent kitchen capacity or build it, and whether you run your own drivers.

Cost Breakdown

• Commercial kitchen or commissary lease/deposit: $5,000-$60,000 (£4K-£45K)
• Kitchen equipment, hot-holding & delivery packaging: $8,000-$70,000 (£6K-£55K)
• Online ordering, POS & kitchen-display integration: $2,000-$18,000 (£1.5K-£14K)
• Delivery fleet, e-bikes/vehicles or driver onboarding: $3,000-$45,000 (£2.5K-£35K)
• Licensing, permits & food-handler certification: $500-$5,000 (£500-£3K)
• General liability + hired/non-owned auto insurance: $2,500-$9,000/yr (£2K-£7K)
• Branding, menu photography & launch marketing: $3,000-$25,000 (£2.5K-£20K)
• Working capital (3 months): $10,000-$40,000 (£8K-£30K)

Funding Routes

In the US, the SBA 7(a) loan covers up to $5M with terms up to 10 years for working capital and equipment, and remains the most common route for food-service operators, the U.S. Small Business Administration reports average 7(a) restaurant loans in the low-to-mid six figures (U.S. Small Business Administration, 2025). Equipment financing and commissary lease-to-own arrangements suit ghost-kitchen buildouts. In the UK, the government-backed Start Up Loan offers up to £25,000 per founder at 6% fixed with free mentoring (GOV.UK Start Up Loans, 2025); many delivery founders stack it with an angel round or a small commercial facility. Our bespoke plan formats the financials to lender and investor expectations, so the funding ask is defensible rather than aspirational.

Planning a wider grocery or convenience delivery angle too? Our grocery delivery business plan template covers the last-mile logistics side in more depth.

Platforms, Software & Suppliers

A delivery operation is a stack of vendors as much as a kitchen. Naming the specific tools in your plan signals to a lender that you have thought past the idea stage. These are the platforms operators actually use:

• DoorDash, largest US marketplace (~65-67% of category sales); tiered commission from roughly 15% up to 30% depending on placement and whether you use its drivers.
• Uber Eats, major US and UK marketplace with similar tiered pricing and a large late-night order base.
• Just Eat / Just Eat Takeaway, dominant UK marketplace, strong in suburban and takeaway-led areas.
• Deliveroo, UK own-logistics network favoured for premium restaurant brands and Editions ghost-kitchen sites.
• Grubhub, US marketplace with a strong presence in the Northeast and campus markets.
• Otter / Deliverect, order-aggregation middleware that pipes all marketplaces into one tablet and your POS, cutting "tablet chaos" errors.
• Toast or Square for Restaurants, POS and kitchen-display systems that unify online, direct, and in-house orders.
• CloudKitchens, Kitchen United & Kitopi, ghost-kitchen infrastructure providers that rent licensed, delivery-ready units by the month.

Packaging suppliers matter more than founders expect: insulated, vented, and tamper-evident containers protect food quality and your review score, and they are a recurring per-order cost your model must carry, not a one-off.

Permits & Legal Requirements

Delivery does not exempt you from food-service regulation, if anything it adds a transport and driver-liability layer on top. Requirements vary by jurisdiction; the essentials are below.

United States

• Food service / health permit from the local county health department ($100-$1,000; 2-8 weeks including inspection).
• Food Protection Manager or Food Handler certification (ServSafe or equivalent) for staff, $15-$180 per person.
• Business licence, EIN and sales tax permit from state, city, and the IRS.
• Commissary or commercial kitchen compliance, most states prohibit delivering higher-risk cooked food from a home kitchen, so a licensed shared kitchen is usually required.
• Hired and non-owned auto or commercial auto insurance for delivery drivers ($1,500-$6,000/yr), plus DOT registration for larger fleets.

United Kingdom

• Register the food business with your local authority Environmental Health team at least 28 days before trading (free, via GOV.UK, 2025).
• Food Hygiene Rating (Scores on the Doors) following a Food Standards Agency inspection.
• Allergen labelling compliance under Natasha's Law / PPDS rules (Food Standards Agency, 2025).
• Level 2 Food Hygiene & Safety training for food handlers (£15-£30 per person).
• Employers' liability, public liability and courier/goods-in-transit insurance for in-house drivers.

Other Jurisdictions

• EU: HACCP food-safety plan under Regulation (EC) 852/2004, national food authority registration, allergen and origin labelling, VAT registration.
• UAE (Dubai): Dubai Municipality food establishment permit, Person-in-Charge certification, cloud-kitchen licences via free zones, third-party liability insurance.
• Australia: state-based Food Safety Supervisor certification, council food business notification, FSANZ compliance, ABN registration, and WorkCover for drivers.

Order Economics & Margins

Delivery revenue is simple to state and easy to get wrong: it is orders multiplied by average order value (AOV), minus the cost of getting each order to the door. US AOV usually sits at $28-$42 (£22-£34 in the UK). The complication is the deduction. On a marketplace order, a blended 15-30% commission comes off before you have paid for a single ingredient. On a direct order you keep that but carry your own delivery cost. That is why the same kitchen can show wildly different margins depending on order mix.

Operators heavily dependent on third-party apps typically net just 3-12%. Own-fleet and ghost-kitchen brands that control commission and drive direct orders reach 12-22%. The lever is not menu price; it is the share of orders that come direct.

A Worked Example

Take a single-brand ghost kitchen doing 60 orders a day at a $34 AOV. That is roughly $61,200 a month, about $734,000 a year. Suppose 70% of orders arrive through apps at a blended 24% commission and 30% are direct at 2% card processing. Effective platform cost is around $115,000 a year. Layer on food cost at 30% ($220K), labour at 24% ($176K), rent and utilities at 9% ($66K), packaging and delivery at 7% ($51K), and marketing at 4% ($29K), and net profit lands near $77,000, roughly 10.5%. Shift the mix so only 55% of orders come through apps, and net margin climbs past 15% on the same sales, the whole thesis of a good delivery plan in one calculation.

Additional revenue layers that strengthen the model: running multiple virtual brands from one kitchen, catering and bulk office orders at higher AOV, branded packaging or merchandise, and subscription or loyalty schemes that lock in repeat purchase.

The Four Levers That Move Net Margin

When we stress-test a delivery model, four variables do almost all the work. Naming them explicitly in your plan shows a lender you understand the mechanics rather than hoping revenue growth alone fixes profitability:

• Order mix (app versus direct): the biggest single lever. Every ten points shifted from app to direct on a $734K-revenue kitchen is worth roughly $18,000 a year in recovered commission at a 24% take rate.
• Average order value: because fixed cost per order (packaging, dispatch, handling) is largely flat, a higher AOV drops more to the bottom line. Meal deals, sides, and drinks are the usual AOV levers.
• Food cost percentage: tight menu engineering and supplier terms keep food cost near 28-32%. A menu of low-cost, high-perceived-value dishes (wings, loaded fries, bowls) is a deliberate margin choice.
• Throughput at peak: the dinner rush is where most orders land, so labour scheduling and prep-ahead systems that hold quality under load protect both margin and rating.

Choosing a Delivery Radius & Kitchen Site

For a dine-in restaurant, location is about footfall. For a delivery brand it is about the population, order density, and drive time inside your radius, a completely different site-selection logic that many plans get wrong by copying restaurant thinking. The kitchen can sit on cheap secondary real estate precisely because no customer visits it; what matters is who lives within a 20-30 minute delivery window.

• Population density: more households per square mile means more orders per driver mile and a lower delivery cost per order. Dense urban and suburban catchments beat sprawling low-density ones.
• Order density by daypart: a catchment with strong dinner and late-night demand supports a comfort-food brand; a business-district catchment favours weekday lunch and catering.
• Drive-time radius, not distance: traffic and one-way systems mean a 3-mile radius in one city delivers faster than 1.5 miles in another. Food quality degrades past roughly 30 minutes, so the radius is a quality constraint, not just a marketing choice.
• Rent-to-revenue ratio: because delivery kitchens do not need prime frontage, target rent well below the 8-10% of revenue a dine-in site would carry, ghost-kitchen units and secondary industrial space are why the model can out-margin traditional restaurants.

The plan should show the estimated addressable orders inside the chosen radius, the competing brands already serving it, and why the site supports the target order volume at an acceptable delivery cost. That analysis is exactly what our Research + Content and Bespoke packages build out with mapped catchment data.

Delivery Margin Calculator

Enter your assumptions to see how order volume, average order value, and the app-versus-direct mix move your annual revenue and rough contribution. This is a planning aid, not a full financial model, the paid packages build the complete Excel forecast.

Figures are indicative and assume roughly 360 trading days. Direct orders are treated as commission-free (a small card-processing cost applies in reality).

Market Size & Demand

US online food delivery revenue reached about $353 billion in 2024, and the meal-delivery segment alone is projected near $390 billion for 2025 (Statista, 2024). Globally the online food delivery market is roughly $1.2 trillion and forecast to grow at around 9% a year (Statista, 2024). This is a delivery-specific figure, not the whole $8T food-and-beverage sector, which is the number that matters for a delivery plan.

Demand is structural, not a pandemic hangover. The National Restaurant Association reports that off-premises dining, delivery, takeout, and drive-thru, now accounts for a majority of restaurant traffic, and around three in four US adults order delivery or takeout at least monthly (National Restaurant Association, 2024). In the UK the online food delivery market is estimated at roughly £12.6 billion, with Just Eat, Deliveroo, and Uber Eats holding most of the share.

The strategic read: the category is enormous and growing, but the marketplaces sit between you and the customer and price that access at 15-30% a transaction. A defensible delivery plan shows how the business builds a direct relationship over time so it is not renting its own customers indefinitely.

Who Actually Orders, And When

A delivery brand lives or dies on the specific customers inside its delivery radius and the moments they order. Generic "everyone who likes food" targeting is why so many delivery plans read as unfundable. The template pushes you to define who the priority customer is, what triggers the order, and why they pick your brand over the twenty others one tap away on the same app.

• Weeknight households: time-poor families and couples ordering dinner between 6pm and 8:30pm. Convenience and consistency matter more than novelty; reliability drives repeat orders.
• Late-night and social occasions: younger customers ordering after 10pm, a segment where Uber Eats and comfort-food virtual brands over-index. Higher impulse, higher AOV, more sensitive to review scores.
• Office and group orders: weekday lunch and catering-style bulk orders that lift AOV well above the solo-diner average and smooth demand across the week.
• Health-led and dietary niches: plant-based, high-protein, or allergen-safe eaters who are underserved by mainstream takeaway menus and reward a brand that speaks directly to them.

Order timing shapes staffing and kitchen throughput as much as it shapes marketing. A brand skewed to a tight 90-minute dinner peak needs different prep and labour scheduling than one spreading orders across lunch, dinner, and late-night. Your plan should quantify each segment's size inside the delivery radius, its average order value, and how the menu and messaging shift by segment, because that is what turns a demand assumption into a credible revenue forecast.

The practical output is a ranked view: which segment produces the best margin, which converts fastest on which app, and which can be reached most cheaply through direct channels once the brand has a foothold. Investors want to see that you know which customer to chase first, not a menu aimed at all of them at once.

Who You Compete Against

Delivery competition is layered, and the layer most founders forget is the one that hurts most: the marketplace-native virtual brands that exist only inside the app and are engineered for the exact search terms customers type. Mapping all three layers is what separates a plan that has thought about competition from one that has merely listed a few local restaurants.

A credible strategy shows how the business wins through tighter execution and a repeat-order relationship, not by underpricing. On the apps, ranking is driven heavily by order volume, ratings, and prep speed, so operational quality is competitive strategy, a brand that consistently ships hot, correct orders at a 4.6 rating out-competes a cheaper rival sitting at 4.1. The plan should map competitor menus, delivery-radius overlap, price points, and where a defensible edge, a signature dish, a diet niche, faster fulfilment, actually exists.

Operations: Where Delivery Margin Is Won

In delivery, operations are not a support function, they are the product. The customer never sees your kitchen, only a bag on their doorstep, so throughput, packaging, and accuracy are the entire brand experience. A strong plan shows exactly how orders are produced, dispatched, measured, and improved as volume climbs.

• Kitchen workflow: prep-ahead systems, station design, and hold times tuned so food leaves at quality even during the dinner rush.
• Order aggregation: piping every marketplace into one screen (via Otter or Deliverect) to kill the "tablet chaos" that causes missed and late orders.
• Dispatch discipline: whether you use app drivers, your own fleet, or a hybrid, with clear rules on when each is used to control cost per delivery.
• Packaging as quality control: vented, insulated, tamper-evident containers matched to each dish so nothing arrives soggy or cold.

Year-One Operating Priorities

• Document every recipe and pack-out spec so quality is repeatable regardless of who is on shift.
• Track owner-level KPIs weekly: average prep time, order accuracy, on-time delivery rate, and marketplace rating by brand.
• Build reporting early so a slipping rating or a rising commission drag is visible before it becomes a structural problem.

The gap between an average delivery operator and a high-performing one usually comes down to four things: prep time under peak load, order accuracy, packaging that holds quality, and the speed at which problems are caught and fixed. Each is a line item your operations plan should own with a target attached.

Marketing: From App Discovery to Owned Customers

The marketing plan should connect each channel to a revenue target and, above all, to the migration from paid app discovery toward owned, direct customers. That migration is the single most valuable marketing outcome for a delivery brand, because it is what pulls commission drag out of the model over time.

• Marketplace optimisation: menu names, photography, and sponsored placement tuned to how customers search on each app, plus ratings management to protect ranking.
• Direct-order channel: an own-branded ordering site, QR codes and inserts in every delivery bag, and a first-order-direct discount to convert app buyers.
• Retention loops: loyalty schemes, SMS/email offers, and lapsed-customer win-backs so acquisition spend compounds instead of resetting each order.

A stronger plan ties these channels to real numbers: customer acquisition cost, conversion rate, repeat-purchase rate, and the percentage of orders shifting from app to direct each quarter. It should show which channel is expected to convert first, what the payback period looks like, and where the founder should concentrate effort before scaling to more brands or a second kitchen.

More Questions Founders Ask

What exactly is a ghost kitchen, and do I need one?

A ghost kitchen (also called a cloud or dark kitchen) is a delivery-only production space with no dining room. Providers such as CloudKitchens, Kitchen United, and Kitopi rent licensed units by the month, so you skip a full restaurant buildout. You do not strictly need one, you can deliver from an existing restaurant or a shared commissary, but for a pure delivery brand a ghost kitchen usually gives the best rent-to-revenue ratio and lets you run several virtual brands from one line.

Can I run more than one brand from a single kitchen?

Yes, and many operators do. Running two or three "virtual brands", say a wings concept, a loaded-fries concept, and a healthy-bowl concept, from one kitchen spreads fixed cost across more orders and lets you test which cuisine converts best on each app. The plan should show shared prep, separate menus, and how you avoid confusing the same customer.

How fast can a delivery brand realistically launch?

From a licensed commissary, a lean marketplace-listed brand can be live in 4-8 weeks once permits and app onboarding clear. A ground-up ghost kitchen with its own fleet is a 3-6 month project. The template includes a launch timeline so your funding request lines up with the ramp.

Do delivery apps share customer data with me?

Largely no, that is the strategic catch. Marketplaces keep the customer relationship, which is why the best plans build a direct-order channel (own site, loyalty, packaging inserts, QR codes) from day one to convert app buyers into owned, repeat customers.

Sample Business Plan Preview

Here is an extract from a delivery-brand business plan written by our team, so you can see the level of specificity a lender or investor expects:

What's in the Template

Every Avvale business plan template includes these sections, pre-structured for a restaurant delivery service and its unit economics:

• Executive Summary, Your delivery brand at a glance, written to hook a lender or investor in 60 seconds.
• Company Overview, Legal structure, kitchen model (marketplace, own-fleet, or ghost kitchen), and founding story.
• Industry Analysis, Delivery market size, marketplace share, and the regulatory picture.
• Customer Analysis, Delivery-radius demographics, order-time patterns, and average order value.
• Competitor Analysis, Local operators, chain delivery, and marketplace-native virtual brands.
• Marketing Plan, App placement, direct-order conversion, loyalty, and repeat-purchase loops.
• Operations Plan, Kitchen workflow, packaging, dispatch, and driver or app logistics.
• Management Team, Founder and kitchen-lead bios, advisers, and key hires planned.

The optional Financial Forecast add-on (included in our $300/£250 and $1,000/£800 packages) provides a 5-year Excel model with income statement, cash flow, balance sheet, break-even analysis, and, specific to delivery, an order-mix sensitivity table that shows how net margin moves as direct orders grow.

Frequently Asked Questions

Related Templates & Resources

Building an adjacent food or logistics business? These Avvale resources pair well with the delivery plan:

• Grocery delivery business plan template, last-mile logistics and basket economics.
• Catering service business plan template, for higher-AOV bulk and event orders.
• Free business plan template, the general-purpose starting structure.
• Market research & content service, done-for-you research and narrative.