Cold Start Discussed
How the first side of the market arrives, before anything is built.
A marketplace has a harder problem than software: buyers will not come without sellers, and sellers will not come without buyers. Marketplace development should start by understanding how you intend to break that, because the technical build is straightforward next to it and cannot solve it.
The most useful thing we can do early is make sure the launch plan is real.
How the first side of the market arrives, before anything is built.
Split payments, holds, refunds and payouts are the hardest part of the build.
Reviews, verification and dispute handling, without inventing signals that do not exist yet.
One category, one city, one segment. Marketplaces launched broadly are thin everywhere.
Listings should be indexable, which affects the architecture — see programmatic SEO.
Four problems, only one of which is software.
Discuss Your Marketplace →Getting one side of the market to arrive before the other exists. The defining problem.
Buyers and sellers who do not know each other, transacting through you.
Taking payment, holding it, taking a fee, paying out, handling disputes.
Both sides have an incentive to transact off-platform after the first match.
Two-sided architecture, the money, and the trust layer between them.
Listing-scale search visibility is covered by programmatic SEO.
Supply plan first, then the narrowest viable market, then the software.
How the first sellers arrive, concretely.
One category or one geography, deep enough to work.
List, discover, transact, review. Nothing else initially.
Payments, fees and payouts, including the failure cases.
Tooling for humans to run the market before automating it.
Rarely by launching a platform and waiting. Usually by faking one side or narrowing until liquidity is achievable.
Narrowing until the market is small enough to fill. A marketplace for everything in a country needs enormous supply; one for a single category in a single city might need a few dozen sellers.
Recruiting supply manually is the other common route. The early sellers are approached individually rather than acquired, which does not scale and does not need to.
What rarely works is building the full platform and marketing to both sides at once. Buyers arrive, find nothing, and do not come back — and the second visit is the expensive one to earn.
Because a marketplace holds funds between two parties. Payment is taken from the buyer, a fee is retained, the balance goes to the seller, and something has to happen when the transaction is disputed or refunded.
That requires a payment provider supporting split payments and connected accounts, plus seller onboarding with identity verification the provider requires.
It is usually the longest single piece of the build and the one most often underestimated, because the happy path is simple and everything around it is not.







Marketplace development builds two-sided platforms where independent buyers and sellers transact — covering listings, discovery, split payments, trust mechanisms and the operational tooling to run the market.
Not the software. It is getting the first side of the market to show up when the other side does not exist yet, which is solved by narrowing the market or recruiting supply manually rather than by building features.
Through a provider supporting split payments and connected accounts: the buyer pays, your fee is retained, the balance is paid to the seller. Seller onboarding includes identity verification the provider requires, which takes longer than expected.
Partly by keeping the valuable parts on it — payment protection, dispute resolution, reviews that matter — and partly by accepting that some leakage happens. Restriction alone tends to push people off faster.
Usually yes, and it affects the architecture. Listing pages can be a significant acquisition channel, provided they are substantial enough to be worth indexing — see programmatic SEO.
Still deciding if marketplace development is right for you?
Talk to UsThe instinct is to launch broadly — more categories, more geography, a bigger addressable market. It feels like ambition and it is usually what kills the launch.
Spread thin, every search returns nothing. A buyer who searches and finds nothing does not merely fail to transact; they conclude the platform is empty and stop returning. That first impression is expensive to reverse.
Marketplaces that work almost always started implausibly narrow — one category, one city, sellers recruited by hand. It looks like a smaller business, and it is the only version that clears.
Tell us what the marketplace connects and how you intend to get the first side. We will be straightforward about whether the plan holds before discussing a build.
