How to validate a marketplace idea before you build.

Marketplaces are structurally different from SaaS. The dominant failure mode is not "the buyer does not exist" (the SaaS pattern) but "the buyer and the seller cannot find each other in enough density to make either side stay". The cold-start problem is the dominant risk, and most marketplace research has to focus on it.

This is a playbook for testing a marketplace idea against the four dimensions that matter most. Each can be tested without writing any code.

Every marketplace starts with a chicken-and-egg problem. Buyers will not come until there is supply; sellers will not list until there are buyers. The marketplace that solves this has identified the side that can be motivated to show up first, and has a credible way to motivate it.

The test: name which side you will recruit first (usually supply for service marketplaces, demand for product marketplaces), and the specific tactic to get the first 50 of them. Concierge supply onboarding, manual outreach, white-glove early sellers, vertical-specific seeding (one city or one category at a time) all work; "we will market it" does not.

For service marketplaces like Upwork in the early days, supply was seeded first by recruiting freelancers manually. For product marketplaces like Etsy, supply was also first but the recruitment pattern was different (community-driven via specific craft niches). The cold-start strategy is specific to the category.

Marketplaces work when buyers and sellers can transact reliably within one market. Reliable means the buyer finds enough supply that the marketplace is worth opening, and the seller finds enough buyers that listing is worth the work.

The test: pick one geography or one vertical and estimate how many sellers you need to make the buyer experience good. For local services, this is usually 20 to 50 sellers in one city. For category marketplaces, it is 100 to 500 sellers in one product category. Then ask: can you realistically recruit that many in the first 90 days?

If the answer is "we will start broad and narrow later", that is the wrong answer. Marketplaces win by starting narrow (one city, one vertical, one niche) and expanding.

Take rate is the percentage of each transaction the marketplace captures. For marketplaces of services or labor (Upwork, Fiverr) it is usually 10% to 20%. For product marketplaces it is usually 3% to 8%. For curated or premium marketplaces it can be higher.

The math: if average transaction is $50, take rate is 10%, and you need 10,000 transactions a month to clear $50K monthly revenue, that is 333 transactions a day across whatever buyer-seller base you have. Can you generate that volume in the geography or vertical you started with? If yes, the unit economics work. If not, either the take rate needs to be higher (premium niche), the transaction value needs to be higher (B2B service marketplace), or the volume math needs to be sourced from a larger initial market.

Marketplaces die when buyers cannot trust sellers or sellers cannot trust buyers. The fix is mechanical, not marketed: reviews, ratings, escrow, dispute resolution, verified identity, money-back guarantees, or some combination.

The test: write down which trust mechanic you will ship at launch. "We will figure out trust later" is the wrong answer. The trust mechanic is part of the product, not a feature that can be added afterward.

The idea: a curated marketplace connecting senior technical writers with SaaS companies that need documentation, API references, and developer-facing content.

- **Cold-start strategy:** supply-side first. Personally recruit the first 50 senior technical writers by reaching out via LinkedIn and existing networks. Pay them a small flat fee for early profile creation if needed. Demand side is recruited after the first 50 sellers are live, via outbound to series A and B SaaS companies known to be hiring tech writers. - **Density:** target the US tech writing market, which has roughly 50,000 active freelance technical writers per BLS data. Recruiting 50 in 90 days is feasible. The buyer side is roughly 10,000 series A through series C SaaS companies, of which a meaningful fraction outsource technical writing. - **Take rate:** 15% on average. Average engagement is $5,000 for a documentation project. 100 transactions a month = $75K take. Realistic at year-two scale. - **Trust mechanics:** seller profiles include verified work samples, client references, and a money-back guarantee for the first engagement. Buyers pay into escrow; payment releases on documented delivery.

All four dimensions clear. This marketplace idea is worth building.

The free Verdikt tests the same four dimensions and returns a Verdikt Score plus the top three named risks. The Single Report ships the full memo with 40+ cited sources, including marketplace-specific competitor analysis and density math grounded in primary data.

The playbook above is the manual version. Verdikt compresses the research timeline and adds external citations to the math.