The question is not which model is better. It is which model fits the supply you can sign and the operation you can run. A multi-vendor marketplace without restaurant density is an empty storefront. A single-brand app without repeat demand is an expensive menu.
| Multi-vendor | Single-vendor | |
|---|---|---|
| Who sells | Many restaurants | One restaurant brand |
| Revenue | Commission + delivery fees | Full basket margin |
| Supply work | Recruit and retain vendors | One menu and kitchen |
| Operations | Vendor support, settlements, disputes | Kitchen and delivery only |
| Software | Vendor onboarding, per-vendor commission, payouts | Single menu, one brand, one zone |
| Best for | A city or category marketplace | A restaurant or small chain |
Multi-vendor is a supply business
Your first job is not software; it is signing 25–40 restaurants in one dense area. Then you need riders, settlement reports, vendor support and a commission that restaurants will accept. The software has to handle per-vendor menus, commissions, zones, self-delivery and payouts. If it cannot, the operation becomes a spreadsheet.
Single-vendor is a brand business
A restaurant app is a direct channel: your menu, your customer list, your offers. The software is simpler — one menu, one kitchen, one delivery rule — but the marketing burden is higher because you do not inherit marketplace demand. The upside is that every repeat order belongs to you, not to an aggregator.
The hybrid model
Many operators start single-brand and add vendors later, or run a multi-vendor platform where one anchor vendor is owned by the platform. Foodomaa supports both: the vendor model is a setting, not a separate product. You can run a single restaurant, a multi-restaurant marketplace, or a mix of platform riders and vendor self-delivery.
What to check in the software
- Vendor onboarding, approval and menu control.
- Commission per vendor, category or order.
- Delivery zones, fees and minimum baskets.
- Self-delivery, platform riders and pickup.
- Settlement reports the vendor can reconcile.
- Source code and data ownership if you plan to customise.
See Foodomaa's multi-restaurant software, the restaurant app, or the launch playbook.
Revenue streams compared
The two models earn money differently. A multi-vendor marketplace usually takes a commission on food sales, a delivery fee from the customer, promoted placement, subscription fees from restaurants and sometimes a small-order fee. A single-brand restaurant keeps the full basket margin but pays for delivery, packaging and marketing itself.
| Revenue stream | Multi-vendor | Single-vendor |
|---|---|---|
| Food margin | Commission only | Full basket margin |
| Delivery fee | Customer or restaurant | Customer |
| Promoted placement | Yes | No marketplace to promote |
| Restaurant subscription | Optional | Not applicable |
| Small-order fee | Common | Possible |
| Packaging / service fee | Optional | Optional |
Commission-only marketplaces often underestimate delivery cost. If the commission is 15% on a $20 order, that is $3 before rider cost, payment fees and refunds. The delivery fee and small-order fee exist to close that gap. A single-brand restaurant has more margin per order but needs enough demand to fill the kitchen.
Pros and cons of each model
Multi-vendor pros: wider selection, higher basket potential, commission revenue, marketplace discovery and the ability to add categories. Cons: restaurant recruitment and retention, vendor support, settlement disputes, quality variance and the cost of building liquidity in every zone.
Single-vendor pros: full control of the menu and customer experience, full basket margin, simpler operations and a direct customer relationship. Cons: you carry the entire marketing burden, limited selection, lower basket size and no commission income.
The hybrid model often wins: one anchor restaurant or dark kitchen owned by the platform, plus independent vendors for selection. The anchor gives you quality control; the marketplace gives you choice.
The feature set that decides
- Vendor onboarding: application, approval, menu upload and training.
- Per-vendor commission: category, order value or delivery type.
- Vendor self-delivery: not every restaurant wants your riders.
- Settlements: order-level statements, refunds, discounts and payout cycles.
- Zones: different radii and fees by vendor or brand.
- Menu control: who can change prices, photos and availability.
- Reporting: revenue, commission, rider cost and contribution by vendor.
If the software cannot separate platform-funded and vendor-funded discounts, finance will spend every payout cycle fixing it. Test that flow before you buy.
A decision scorecard
| Question | Multi-vendor | Single-vendor |
|---|---|---|
| Can we sign 25–40 restaurants in one zone? | Required | Not needed |
| Do we have rider operations experience? | Important | Important |
| Is our brand strong enough to drive demand? | Less critical | Critical |
| Do we want commission income? | Yes | No |
| Can we support vendors and disputes? | Yes | Not applicable |
| Is the menu standard or novel? | Standard fits | Novel fits |
Common launch mistakes
- Launching a marketplace with 10 restaurants spread across a city.
- Setting commission without modelling rider cost and payment fees.
- Letting vendors self-deliver without a service standard.
- Paying restaurants late or without a reconcilable statement.
- Adding a second zone before the first one is profitable.
- Choosing a single-vendor model without a repeat-demand plan.
Foodomaa supports both models on one installation, so the decision is reversible. Start with the model that matches the supply and demand you can prove in one zone, not the one that sounds bigger.
A unit-economics example
Take a $25 order. Multi-vendor: 15% commission ($3.75) + $2 delivery fee − $2.50 rider − $0.63 payment fee = $2.62 contribution. Single-vendor: $25 food revenue − $8 food cost − $2.50 rider − $0.63 payment − $0.50 packaging = $13.37 contribution, but the restaurant also carries marketing and kitchen overhead.
The single-vendor number looks better per order, but it depends on demand. A marketplace with 1,000 orders and $2.62 contribution makes $2,620 before fixed costs. A single restaurant with 300 orders and $13.37 contribution makes $4,011. The marketplace has more scale potential; the restaurant has more margin per order.
How to recruit 30 restaurants
- List every takeaway-capable restaurant in the zone, not just the popular ones.
- Lead with a lower commission or a free first-month trial, not a vague partnership pitch.
- Bring a tablet, a menu template and a photographer; reduce the vendor's work.
- Sign 25–40 before launch so the marketplace does not look empty.
- Track menu completeness, prep time and response time before adding more vendors.
Vendor retention after month one
Vendors leave when orders are low, payouts are late or support is slow. Publish a payout calendar, send a statement with every payment and give each vendor a named contact. Review rejection rate and prep-time accuracy monthly. A vendor who feels like a partner will stay through a slow month; a vendor who feels like a line item will leave at the first better offer.
When to switch models
Start single-vendor if you own the kitchen or have one anchor brand. Add vendors when demand exceeds your menu capacity. Start multi-vendor if you can sign 25–40 restaurants and support settlements. Switch from single to hybrid before switching to a full marketplace; the hybrid lets you test vendor onboarding without losing quality control.
Liquidity math: orders per restaurant per day
A marketplace is liquid when a customer can find something they want and a restaurant receives enough orders to stay. The simplest measure is orders per active restaurant per day. Below five, most restaurants treat the platform as an afterthought and stop updating menus. Above fifteen, the best vendors start to depend on you and will answer the tablet.
Take a zone with 30 restaurants and 300 orders a day: ten orders per restaurant. That is workable, but only if the orders are spread sensibly. If the top five restaurants take half the volume, the other 25 are below three orders a day and will churn. Track the distribution, not just the average. A marketplace with a healthy average and a long tail of inactive vendors is not liquid.
Use this to size the launch zone. If you can recruit 25 restaurants but only generate 100 orders a day, the average is four. Either shrink the zone, increase demand, or wait. Adding more restaurants makes the ratio worse. The number is a hard guardrail against the most common launch mistake: supply without demand.
Commission and fee architecture
Commission is only one line in the vendor's payout. The rest of the architecture decides whether vendors trust the statement:
| Line | Who pays | Common mistake |
|---|---|---|
| Food commission | Vendor | Changing the rate without notice |
| Delivery fee | Customer, vendor or platform | Not showing who keeps it |
| Payment fee | Platform or vendor | Double-charging the vendor |
| Promoted placement | Vendor | Mixing ad spend with commission |
| Refund | Depends on fault | Always deducting from the vendor |
| Small-order fee | Customer | Surprising the customer at checkout |
| Tax | Legal entity | Treating tax as revenue |
Publish a sample payout for a $25 order. Show gross food sales, commission, delivery, payment fee, refund reserve and net payable. A vendor who can reproduce the number will argue less and stay longer. If the statement only shows a final figure, every dispute becomes a support ticket.
Keep the fee structure simple at launch. One commission rate, one delivery fee rule and one payout cycle. Add complexity only when the operation demands it. A marketplace with five fee types on day one is hard to explain and harder to reconcile.
Vendor contract and onboarding terms
The contract should answer the questions that cause disputes later: menu price parity, delivery responsibility, packaging standards, cancellation windows, refund liability, payout timing, exclusivity, data ownership and notice period. Verbal promises at recruitment become expensive exceptions.
Put the operational terms in an onboarding checklist the vendor signs: prep time, opening hours, pause rules, photography approval, tablet response time and who handles a missing item. Then enforce them consistently. A vendor that ignores orders for 20 minutes should be paused, not given a special case, because the customer blames the platform.
Do not sign long exclusivity at launch. You do not have the demand to justify it, and it makes recruitment hostile. Offer a lower commission for a defined launch period instead, with a clear end date. If the vendor stays after the period, the normal rate applies.
Dispatch model and vendor type
Not every vendor should use the same delivery model. A high-volume restaurant inside a dense zone may work well with platform riders. A bakery with its own driver may be cheaper on self-delivery. A grocery store may need batch picking and a different vehicle. Configure by vendor, not by platform.
If vendors self-deliver, the platform still owns the customer promise. Track their pickup and delivery times, and remove the option if they repeatedly miss the estimate. If platform riders serve every vendor, zone density and rider shift planning become the core operating job. Choose the model that matches your supply, then measure the handoff.
The cold-start plan
A marketplace with no demand cannot recruit supply, and a marketplace with no supply cannot attract demand. Break the loop with a narrow launch, not a city-wide map.
- Weeks 1–2: pick one 3–4 km zone, list every takeaway restaurant and sign 20–30 before announcing.
- Weeks 3–4: onboard menus, photos and prep times; run test orders with staff and friends.
- Weeks 5–6: launch to one customer segment — a campus, office park or residential complex — with a simple first-order offer.
- Weeks 7–8: measure orders per restaurant, repeat rate, delivery time and refunds; fix the weakest number before adding a second zone.
Do not measure success by registrations. Measure by repeat orders and vendor acceptance rate. A launch with 500 sign-ups and 30 repeat customers is a failed launch; 80 customers ordering twice a week is a business.
Keep the launch offer simple and time-boxed. A free-delivery week is easy to understand; a complex coupon stack is hard to reconcile and attracts one-time bargain hunters. Once the zone is liquid, reduce the subsidy and watch the repeat rate. If it collapses, the product was the discount, not the marketplace.
A vendor scorecard
Review vendors monthly on four numbers: order acceptance rate, average prep time, cancellation rate and menu freshness. A vendor below 90% acceptance or above 25 minutes prep is hurting the customer promise, even if sales are strong. Put the score in the vendor statement and use it in the quarterly review.
Do not punish a vendor for a platform failure. If riders arrive late, prep time is not the vendor's fault. Separate the metrics you control from the ones they control, and publish the definitions. A scorecard without definitions becomes a negotiation, not an operating tool.