A marketplace app connects buyers and sellers and takes a cut or a fee. The first release needs listings, search, a way to transact and a way to trust the other side. Everything else, including messaging, reviews, disputes and a second platform, can come later. A focused marketplace MVP typically costs $60,000 to $150,000 (€50,000 to €130,000, £44,000 to £110,000) and takes 14 to 20 weeks. The hardest part is not the software, it is getting supply and demand to show up at the same time.
Prices are in US dollars. Euro and pound figures in parentheses are approximate, rounded conversions. Quotes are issued in the currency of your contract.
Marketplaces are among the most attractive businesses to found and among the hardest to launch. The software is only half of it. This guide covers what to build first, the decisions that shape the product, the payments question that trips up most founders, and a realistic budget.
What kind of marketplace are you building?
The answer decides most of the scope, so settle it first:
- Product or service? Products need inventory, shipping and returns. Services need scheduling, location and often quotes.
- One-off or recurring? A wedding photographer is booked once. A dog walker is booked weekly. Recurring changes payments and retention.
- Managed or open? A managed marketplace vets supply and handles quality. An open one lets anyone list. Managed is more work per seller and far easier to launch with quality.
- Local or global? Local means location search and a city-by-city launch. Global means currencies, languages and tax.
Which features belong in the first release?
| Area | First release | Later |
|---|---|---|
| Supply side | Seller sign-up, profile, create and manage a listing | Analytics dashboard, promotions, bulk tools, seller app |
| Demand side | Browse, search and filter, view a listing, contact or book | Saved searches, favourites, recommendations |
| Transaction | Book or buy, pay, confirmation | Subscriptions, instalments, gift cards, promo codes |
| Trust | Verified sellers, clear policies, a visible support channel | Ratings and reviews, identity verification, insurance |
| Communication | Email and notifications on key events | In-app messaging, calls, negotiation |
| Operations | Admin panel to approve listings, see transactions and refund | Dispute workflows, fraud tools, automated payouts |
| Platforms | Responsive web, or web for sellers and one mobile app for buyers | Native apps for both sides |
The pattern to notice: the first release is a complete loop, from listing to payment, with trust supplied by you rather than by software. Reviews are the feature founders most want at launch and the one that is useless until there are transactions to review.
How do you handle payments?
Marketplaces are harder than ordinary e-commerce because money comes in from a buyer and goes out to a seller, with your fee in between. You have three choices:
- Off-platform. Buyers and sellers settle directly and you charge sellers a listing or subscription fee. Simplest to build, and you lose visibility and most of the revenue model.
- A marketplace payment provider. Services such as Stripe Connect or similar handle collecting payment, holding it, splitting your fee and paying out sellers, along with seller identity checks and much of the regulatory burden. This is what almost every marketplace should do. It adds two to four weeks to a build and removes months of compliance work.
- Your own payment handling. Holding client money yourself makes you a payment institution in most jurisdictions, with licensing to match. Not for a first release, and rarely ever.
Decide early when sellers get paid: on booking, on completion, or after a dispute window. That single decision shapes the payment flow, the cancellation policy and the support load.
What is the chicken-and-egg problem and how do you solve it?
Buyers will not come without sellers and sellers will not list without buyers. Every successful marketplace solved this by hand before the software did it for them:
- Start on one side. Usually supply. Recruit the first fifty sellers personally, before launch, and make their listings excellent.
- Start in one place. One city or one niche, where fifty sellers is a real choice for buyers.
- Be the demand at first. Bring buyers through your own marketing, partnerships and, if needed, by fulfilling early orders yourself.
- Make the single-player mode useful. Give sellers a reason to use the product even before buyers arrive, such as a booking page or invoicing.
Our post on validation before code describes running the concierge version of a marketplace, which is the best way to learn how yours will work.
What technical decisions matter most?
- One backend, multiple front ends. Sellers and buyers are different users, often on different devices. A single API serving a web app and a mobile app keeps the data consistent.
- Search. Filtering by location, availability, price and category is the core of the buyer experience. Build it with a proper search service from the start.
- Data model for listings. A listing for a product, a service, a rental and an event are different. Decide what yours is and resist making it generic.
- Roles and permissions. Buyer, seller, admin, and often a user who is both. Get this right in the first release or everything after is harder.
What does it cost and how long does it take?
| Scope | Typical cost | Timeline |
|---|---|---|
| Single-sided listing site with contact, no payments | $25,000 to $50,000 (€22,000 to €44,000, £18,000 to £37,000) | 8 to 10 weeks |
| Two-sided web marketplace with payments and payouts | $60,000 to $150,000 (€50,000 to €130,000, £44,000 to £110,000) | 14 to 20 weeks |
| Two-sided with web and mobile, messaging, reviews, disputes | $150,000 to $350,000 (€130,000 to €305,000, £110,000 to £260,000) | 6 to 9 months |
See what an app costs in 2026 for what drives these numbers.
What mistakes do marketplace founders make?
- Building both sides fully before either has users.
- Launching everywhere instead of somewhere.
- Handling money themselves.
- Adding reviews, messaging and a referral scheme before the first hundred transactions.
- Letting sellers take transactions off-platform because the on-platform experience adds friction without adding value. Give both sides a reason to stay: protection, convenience, records.
How 7L builds marketplaces
We have built two-sided platforms for e-commerce, hospitality and services, and we start every one by defining the first loop and the launch niche before a screen is designed. Payments go through a marketplace provider in your name. The first release is scoped to a fixed budget and timeline. If you are planning a marketplace, tell us about the two sides and we will help you decide which to build first.
Frequently asked questions
Should I build a marketplace with no-code first?
Often, yes. Listing, search and simple booking are within reach of no-code tools, and a no-code marketplace with real transactions is excellent evidence. Payments with payouts and custom matching are where the tools run out. See our no-code versus custom guide.
How much commission should a marketplace charge?
It varies by category from a few percent for high-value goods to 20 percent or more for services where the platform brings most of the demand. Set it from what the platform adds, and test it. It is easier to lower a fee than raise one.
Do I need a mobile app for a marketplace?
Not for a first release, in most cases. Sellers manage listings on the web. Buyers browse on the web, including on phones. Build a mobile app once you know which side uses it most and why.
How do I prevent buyers and sellers dealing off-platform?
By making the platform worth using: payment protection, dispute handling, records, convenience. Rules alone do not work. If users leave, the product is missing a reason to stay.
What about a marketplace where I am the only seller?
That is an e-commerce or booking platform, not a marketplace, and it is considerably simpler to build. See our guides to building an e-commerce platform and a booking platform.