Food Delivery Business From Home

Starting a Ghost Kitchen Is Mostly About Logistics, Not Cooking

I started a food delivery operation out of my apartment kitchen three years ago. I thought the hard part was making good food. It wasn't. The hard part was realizing that every Uber Eats order came with a 25-30% platform cut, a driver who couldn't find my building's back entrance, and customers who expected gourmet results at fast-food prices. Let me walk you through how this actually works in practice, not the version you see on YouTube. The model itself is simple enough. You prepare food in a home kitchen, list it on delivery platforms, and someone picks it up or you arrange drop-off. But the devil is in the operational details that nobody talks about until you're already losing money.

Setting Up a Food Delivery Business From Home

You need three things before you take your first order: a legal setup, a menu that can actually survive the delivery process, and a kitchen that passes inspection. Most people skip step one and get their ass handed to them by local health departments. I've seen people shut down after two months because they didn't know their city requires a commercial-grade grease trap even for home-based operations. Call your local health department first. Don't research it online. Just call them and ask what you specifically need. For licensing, most jurisdictions have a cottage food law that allows certain non-perishable items to be made at home without a full commercial kitchen inspection. That means cookies, jams, bread. If you're doing actual prepared meals with proteins and dairy, you typically need a licensed commercial kitchen or at minimum a permitted home kitchen inspection. The cost varies wildly by location. In my city, a home kitchen permit ran about $800 upfront plus $200 annually. That's the cheap end. In some cities it's three times that. Menu design is where most beginners torpedo themselves. Your menu needs to survive 20-30 minutes in a plastic bag on a moped seat. Fried foods are a death sentence for texture. Soups spill. Anything with a complex assembly requirement will make your prep times explode during peak hours. I learned this the hard way. My third week, I offered a deconstructed tacos plate that looked beautiful in photos. By the time it reached the customer, the tortilla was soggy, the lettuce had wilted into a sad green paste, and the sour cream had separated into an oily mess. One star review with a photo. I took that item off the menu immediately and replaced it with a burrito. Burritos travel fine. The margins are decent too, which is the part that matters.

The Platform Decision Changes Everything

You have a few paths for getting orders. Third-party aggregators like Uber Eats, DoorDash, and Grubhub are the easiest to start with but the most expensive. They take between 15% and 30% per order depending on your contract and whether you handle delivery yourself or let them handle it. For a beginner, this margin compression is brutal. A $15 bowl that looks profitable on paper might net you $9 after fees and payment processing. At volume, it adds up to a loss. The alternative is building your own ordering system. This means a simple website with an ordering interface, your own delivery logistics or a local courier arrangement, and payment processing. The margins are dramatically better — you keep nearly all of it. The catch is that you now have to do your own marketing, handle customer service directly, and manage delivery routing. My first month using DoorDash, I made $2,400 in gross revenue and took home about $1,100 after all fees, tips not included. Month four, after switching to my own ordering system with a small fleet of contracted drivers, gross was $2,800 and net was $2,100. Same volume, different math. Here's a counter-intuitive thing most guides won't tell you: third-party platforms can actually be useful for new operations even though they eat your margins. The reason is discovery. Getting your first 100 orders through your own website is extremely difficult. Getting them through DoorDash is just a matter of setting up your listing correctly and having decent photos. Use the platforms as a customer acquisition channel for the first 90 days, then aggressively push customers toward your direct ordering system with insert cards, SMS flows, and loyalty discounts.

Operational Realities Nobody Warns You About

Packaging costs will eat you alive if you don't plan for them. I budgeted $0.45 per container. Actual cost landed at $1.10 when I factored in lids, napkins, utensils, sauce packets, and the bag itself. That single line item difference cost me roughly $400 a month at my volume. Start with unit economics on every single component before you launch. Write it down. I use a spreadsheet that tracks ingredient cost, packaging cost, labor time, and platform fee for each menu item. If the margin isn't at least 60% after everything, the item doesn't go on the menu. I've seen plenty of people skip this step and wonder why they're working 14-hour days and making less than minimum wage. Peak hours are brutal and they happen at the same time everywhere. 12 to 1 PM on weekdays, 5:30 to 8:30 PM every day. Your kitchen needs to be able to handle three to five simultaneous orders without the quality dropping. This means prepping ingredients in bulk ahead of time, having a strict order of operations, and knowing exactly how many orders your setup can handle before you start turning business away. I capped myself at 25 orders per shift for the first six weeks. Going past that without systems in place means mistakes, cold food, and angry customers. The algorithm punishes you for bad orders by dropping your visibility. It's a vicious cycle. Here's another specific problem I ran into that I wish someone had told me about: delivery radius miscalculation. The platforms show you the radius but don't always account for traffic patterns, road closures, or the fact that some neighborhoods are much harder to reach than others. I had a delivery to a strip mall that took 22 minutes one way during a 15-minute promised window. The customer got cold food, complained, and I ate the refund plus lost the delivery fee. Now I map out every delivery zone manually and build in buffer time. I also avoid certain zip codes entirely during rush hour because the traffic pattern makes them unreliable. This isn't in any guide. You learn it the painful way.

When This Model Doesn't Work

Food delivery from home is not a side hustle. It's a real business that demands real hours, especially in the beginning. If you're planning to do this alongside a full-time job and expect to make passive income, stop reading now. The first three months require 50-60 hour weeks minimum. After that it might settle to 35-40 hours if your systems are solid. It also doesn't work well in markets with low average order values. If the typical order in your area is $12 and below, the platform fees and packaging costs make it nearly impossible to be profitable unless you have very high volume. I'd recommend a minimum average order value of $18 before you start. Check the local competition on the apps. Look at what similar restaurants are charging. If everyone's discounting heavily to $10-12, you're in a tough market. Another scenario where this fails: areas with strict enforcement of home kitchen cooking laws. Some cities actively patrol for unlicensed food operations, especially after a complaint. I know someone in Ohio who got a cease and desist after a neighbor reported the smell of cooking coming from their garage apartment. They lost six months of revenue and had to scramble to find a commissary kitchen. Research your city's enforcement posture before investing anything.

Getting Started Step by Step

Week one: call your health department, understand the requirements, and get the paperwork started. This alone can take 2-4 weeks depending on your jurisdiction. Week two: design a lean menu with 6-8 items maximum. Each item should be able to be assembled in under five minutes during peak service. Week three: source packaging in bulk from restaurants supply companies, not Amazon. The per-unit cost difference is significant. Week four: set up your kitchen, do test runs, time everything, and refine your process. Week five: launch on one platform, maybe two. Don't try to be everywhere at once. Week six onward: start building your direct ordering system and gradually shift customers over. I used Square for my direct ordering initially because it was free to set up and handled payments reliably. Later I moved to a dedicated restaurant POS system once volume justified the $50-100 monthly cost. The integration with inventory tracking alone was worth it. You'll know exactly which ingredients you're running low on before you run out, which prevents the nightmare of taking orders you can't fulfill. The food delivery space is crowded and the margins are tight, but it's viable if you treat it like a real operations challenge rather than a cooking passion project. The people who succeed aren't necessarily the best cooks. They're the ones who understand their unit economics, manage their throughput efficiently, and build a brand that customers will order from repeatedly. Start small, track everything, and scale only when the numbers prove it's sustainable.