Baltimore's independent restaurants — the crab houses, the market stalls, the rowhouse corner spots that give this city its table — mostly take online orders one of two ways: through a third-party marketplace that skims 15 to 30 percent off the top, or through a commission-free tool like ChowNow, Toast or Square that charges a flat monthly rent instead. Leaning less on the marketplaces is almost always the right move, and a commission-free tool is the right next step for most kitchens — we'll say so plainly. But some restaurants outgrow the rented template: several locations, a real catering or wholesale side, a meal-prep or membership model the software can't quite hold. This is an honest guide to that line — what the marketplaces really cost, where the flat-rate tools stop fitting, and when a custom online ordering system you own outright is worth building in Baltimore in 2026.
Baltimore eats out on its independents
You can read a city by where it eats, and Baltimore reads as an independent town. The chains are here like they're everywhere, but the places people actually name are the ones with a single roof and an owner you can find in the kitchen — the crab houses down at the water, the corner spot that's been doing the same coddies for forty years, the Fells Point oyster bar that landed on a national best-of list last year. Lexington Market, the longest continuously operating public market in the country, reopened in 2023 in a new hall after more than two centuries of trading, and its most famous tenant, Faidley's, has been selling crab cakes across the counter since long before anyone said the word "software." Broadway Market, Highlandtown, Hampden, Little Italy, Federal Hill — the map of this city is a map of small food businesses.
Every one of them runs on software now, whether the owner frames it that way or not. The menu, the online orders, the pickup queue, the delivery hand-off, the quiet nightly question of who ordered what and whether you'll ever hear from them again — all of it lives in an app. And here's the part that gets skipped: which app you take orders through, and how much it skims on the way past, quietly decides whether a busy Friday actually leaves anything in the till. On a plate with a fifteen-cent margin per dollar, a platform that takes thirty cents of every online dollar isn't a fee. It's the difference between a good night and a break-even one. That's a big enough lever to be deliberate about, rather than defaulting into whatever the delivery rep signed you up for two years ago.
The three ways a restaurant takes an order online
Strip away the logos and there are really only three. The first is the third-party marketplace — DoorDash, Uber Eats, Grubhub — where a customer finds you inside their app, orders, and the platform handles the driver and takes a cut of the whole ticket. The second is a commission-free ordering tool — ChowNow, Toast, Square, Popmenu and the rest — that puts an ordering button on your own website and charges you a flat monthly fee instead of a percentage. The third is a custom online ordering system you have built and then own outright: your own branded storefront, your menu and rules, your customer list, your code.
Most Baltimore restaurants live somewhere in the first two, often in both at once — on the marketplaces for reach and on a flat-rate tool for their own direct orders. The third option barely gets discussed, because for years it meant an agency, half a year and a number with too many zeros. That math has changed, and the rest of this piece is about when the third option is now the right one. But it only makes sense once you're honest about what the first two actually cost.
What the marketplaces actually cost
Start with the marketplaces, because that's where the real money leaks. In 2026 the commission bands are roughly where they've been: DoorDash runs three tiers at 15, 25 and 30 percent, Uber Eats sits around 20 to 30 percent depending on the plan, and Grubhub's marketing commissions run 5 to 20 percent before you add its delivery fee. Those are the headline numbers, and they're bad enough. The trouble is that the headline rarely holds. Once you factor in the promoted-placement fees you buy to stay visible, the promotional discounts the platform nudges you into, and the pile of service add-ons, restaurants routinely report an effective cost of 30 to 40 percent per order. On thin restaurant margins, an order that leaves through a marketplace can be a wash or a small loss dressed up as revenue.
And the commission is only the visible cost. The one that compounds is quieter: the marketplace owns the customer, not you. The person who ordered your crab cakes is a DoorDash customer who happened to pick your restaurant that night. You don't get their email, you can't invite them back, you can't tell them about the new catering menu, and the app is free to show them a competitor next time. You're renting reach by the order and handing over the one asset — the relationship — that would let you stop renting it. Nobody in the industry seriously argues you should be marketplace-only; the honest debate starts one step past that, at how you take your own orders.
Commission-free tools are the right first move for most
This is the step almost every restaurant should take, and I want to be as plain about it as I'd be on a call: for most kitchens, the answer is a commission-free ordering tool, and you do not need us to build you anything. Get an ordering button on your own site, keep the customer, pay a predictable flat fee instead of a percentage, and get on with cooking. A studio that only ever tells you to build custom is selling you something you probably don't need.
The options are mature and genuinely good. ChowNow charges a flat monthly fee — roughly $119 to $328 a month depending on the tier — takes no commission on your direct orders, and reckons it saves the average restaurant something like sixteen thousand dollars a year against the marketplaces. If you already run Toast as your point of sale, its online ordering can add little on top of what you're already paying, and it keeps everything in one system — though the real Toast bill lives in the hardware, the add-ons and the card processing. Square is the low-barrier entry point, with a free tier and simple per-transaction pricing. For a single-location restaurant with a normal takeout and delivery menu, one of these will fit like a glove, and I'd tell you to buy it and never think about it again.
So where's the catch? There isn't one, exactly — there's a ceiling. These tools are still rented, still monthly, still built to a template that assumes you sell the way ten thousand other restaurants sell. You pay per location and per add-on, so the number climbs as you grow. And your customer list, your order history, your whole operating record still live inside someone else's database, exportable on their terms and shaped by their roadmap. For most restaurants that ceiling is high enough that you'll never touch your head on it. For a few, it turns out to be lower than the business needs.
The day the template stops fitting
You feel that ceiling in specific, recognizable ways, and they rhyme with what any growing business feels when it outgrows off-the-shelf software. The clearest one is locations. The moment you're running two, three, five kitchens, you want shared menus with per-location pricing, one customer list across all of them, one loyalty balance a regular can spend anywhere, and reporting that lets you compare Tuesday in Canton against Tuesday in Towson. Flat-rate tools bolt multi-location on, but it often means paying per site for something that still treats each one as an island, and you end up with a manager stitching the real picture together in a spreadsheet every Monday.
Catering and wholesale are the next. A takeout template understands a cart and a checkout; it does not understand a corporate lunch order placed nine days out with a deposit, a headcount that changes twice, a delivery window, an invoice and net-30 terms. The same is true of a real wholesale side selling to shops and offices. That's often the highest-margin revenue a restaurant has, and it's the part the ordering software handles worst, so it quietly runs on email, a PDF and someone's memory. Then there are the models the template simply doesn't have a box for at all — a meal-prep subscription that bills weekly and lets a customer skip a week, a membership that changes every price at checkout, a bakery pre-order calendar, a ghost-kitchen setup running three brands out of one line. None of this is exotic. It's just a business that has grown its own way of selling, and every workaround is a small tax you pay in perpetuity.
The tell is the same one it always is. When you start paying people, or burning your best people's hours, to serve the software instead of the software serving them — when the catering manager spends half her week being the integration nobody built — the calculus has changed. That's the moment worth stopping to do the math.
What "cheap" ordering software costs over three years
The reason restaurants tolerate the workarounds is that each monthly number looks small next to a payroll and a food bill. But run it out honestly. A flat-rate tool at a couple hundred dollars a month is a few thousand a year for one location; multiply by locations, add the loyalty add-on and the catering add-on and the reservations add-on, layer card processing on every order, and a growing multi-location operation is quietly spending real five-figure money a year on a rented stack that still doesn't hold its catering. At the end of three or five years of that, you own exactly nothing — and your entire customer and order history is sitting in databases you don't control, which is the one thing you'd most want to keep if you ever left.
That is not an argument against renting. Plenty of restaurants should rent forever and never give it another thought, and I've told owners exactly that. It's an argument for knowing the true multi-year number, across every location and add-on, before you decide the daily workarounds are cheaper than the alternative. Usually nobody has actually added it up. When you do, the comparison to a one-time price you'd own outright sometimes looks very different than the small monthly figure suggested.
What a custom online ordering system costs to build in Baltimore
Here's where owning stops being a luxury and turns back into arithmetic. The old objection was that custom software cost a fortune and took most of a year, and that was true when it meant an agency renting you a team of five by the hour. It isn't true anymore. When a senior builder has AI carrying the repetitive eighty percent of the work, the same tools that build venture-backed startups can be pointed at a three-location Baltimore restaurant on a timeline and at a price that make sense for one. We freeze the scope and put a fixed number on it before any code is written — a real number you agree to up front, not an hourly estimate that drifts through the summer. For a restaurant it usually looks like one of these.
| What you're building | What it is | Fixed price | Timeline |
|---|---|---|---|
| Prototype Sprint | One flow — your online ordering page, clickable and deployed | $3,500 | ~1 week |
| Online Store | Branded ordering storefront — menu, modifiers, cart, Stripe checkout | from $6,000 | 1–2 weeks |
| Custom App | The ordering, catering or loyalty app your kitchen runs on | from $12,000 | 2–4 weeks |
| Operations System | Orders, kitchen, inventory and reporting across locations | from $12,000 | 2–5 weeks |
The honest way to start is small. A one-week Prototype Sprint puts a clickable version of your actual ordering flow in front of you for $3,500 — before you commit to anything, you get to see your own menu, modifiers and checkout running as software and decide for yourself whether it beats what you have. From there, a custom Online Store (from $6,000) is the core of it: a branded ordering storefront with your full menu, item modifiers and combos, a cart, and Stripe checkout for pickup and delivery — a real online shop for your kitchen, deployed and fully yours, not a widget bolted onto someone else's platform. It's the same kind of build we broke down step by step in how we built a full e-commerce store in a weekend. When the ordering is only part of it — catering with deposits and lead times, a loyalty program on your own rules, a meal-prep subscription — a custom app or operations system (from $12,000) runs the whole slice end to end: an order comes in, gets routed to the right kitchen, gets made, gets fulfilled, and lands in a report you actually trust, across every location, in one place you own.
The difference that matters isn't the sticker price — it's what you own underneath it. You own every line of the code, the keys, the accounts, and, crucially, the customer list and the order history. There's no per-order commission, no per-location seat, no add-on that appears on next month's bill, and no template you have to bend your catering around because a vendor in California decided years ago how a restaurant ought to sell. Open your fourth location and the software costs what it did for your third. We took a typical build apart line by line in what a custom app really costs, and if you're wondering how the timeline collapsed from months into weeks, the longer answer is in how long it takes to build a custom app.
So — rent or own?
You don't need a consultant to make this call. You need to answer three questions honestly. First: are you still leaning hard on the marketplaces for your own repeat customers? If so, that's the leak to fix before anything else — move those regulars onto your own ordering, whatever tool you use, and keep the relationship. Second: does a flat-rate ordering tool actually fit the way you sell, or are you already paying people and patching spreadsheets to work around it — the multi-location picture, the catering that runs on email, the subscription the template can't bill? If it fits, keep renting; you will not beat a mature product on its home turf, and you shouldn't try. Third: what is the rented stack really costing you across every location and add-on in a year, and what will it cost at twice your size? Line those three numbers up next to a one-time fixed price you'd own forever, and the answer is usually obvious within a minute.
Most Baltimore restaurants should get off the marketplaces for their direct orders, run a good commission-free tool, and get back to the pass — and we'll happily be the ones to tell them so, with no invoice attached. The ones that should build tend to already sense it: they can feel the software fighting the way they actually make money. For those, owning the system they sell through outright is one of the better investments they'll make this decade. If you're genuinely unsure which side of the line you're on, we wrote a broader framework for exactly that decision in custom build vs SaaS vs no-code, and our Baltimore custom-software guide covers the wider local landscape beyond restaurants.
Built in Baltimore, yours to keep
We're a small studio of ex-founders based right here in Baltimore, and we build custom software for small and growing businesses — the online stores, ordering apps, internal tools and operations systems that a team actually runs on. Fixed price, fixed timeline, direct with the builders, and fully yours at the end. We work with clients across the US and Europe, so being outside the city is no obstacle at all, but there's something we genuinely like about building software for the kitchens that feed our own town.
If your restaurant has outgrown the ordering tool it runs on, the way to find out what a custom one would take is a free thirty-minute call. Bring the part that never fit — the catering that lives in your inbox, the second location that doesn't talk to the first, the subscription you can't bill — and we'll tell you honestly whether you should build it or stay put, how fast it could ship, and the fixed price that goes with it. If you'd rather see custom software work before you talk to anyone, we keep five real apps running live in the browser on the demos page, including a full e-commerce store.
Common questions from Baltimore restaurants
Is a custom online ordering system worth it for a small restaurant?
For most single-location restaurants, no — a commission-free tool like ChowNow, Toast or Square is the right buy, and you should use it. A custom online ordering system starts to pay off when the template stops fitting how you actually sell: multiple locations with shared menus, a real catering or wholesale side with deposits and lead times, a meal-prep or membership model, or a stack that has sprawled into several rented tools that each take their own fee. At that point owning one system built around your business usually beats renting three that don't quite fit.
How much does a custom online ordering system cost compared to ChowNow or Toast?
A custom build is a one-time fixed price you own; ChowNow and Toast are rent you pay every month, forever. Our packages start at a $3,500 one-week prototype of your ordering flow and a branded custom online store from $6,000, with a full ordering, catering or operations app from $12,000. By contrast, ChowNow runs roughly $119 to $328 a month, and Toast, Square and the rest layer monthly fees, per-location seats and card processing on top — every month, across every location, rising as you grow.
Can a custom system replace DoorDash, Uber Eats and Toast online ordering?
Yes. A custom online ordering system can take pickup, delivery, catering and subscription orders directly from your own branded site, built around your real menu, modifiers and pricing instead of a generic template. You keep 100% of the order value minus card processing, and you own the customer list, the order history and the code outright — none of which you get from a marketplace that takes 15 to 30 percent and keeps the customer for itself.
How long does it take to build a custom online ordering system?
Weeks, not months. A clickable prototype of your ordering flow takes about a week; a branded online store with menu, cart and Stripe checkout usually ships in one to two weeks, and a full ordering, catering or operations app in two to four. The old multi-month agency timeline was mostly coordination overhead — with a senior builder and AI carrying the repetitive work, that is exactly what disappears.