How to Validate a Marketplace Idea Before You Build Anything

Every week I talk to a founder who has spent six months and $50,000 building a marketplace that nobody uses. The idea sounded great. Their friends thought it was brilliant. They had 200 people on a waitlist. And then they launched, and nothing happened.

The problem was not the execution. The problem was that they validated the wrong things before they built.

What does not count as validation

Before getting to what works, it is worth being explicit about what does not:

• Surveys. People say what they think you want to hear. "Would you use a platform that connected X with Y?" gets a yes from nearly everyone. It tells you nothing about whether they would actually use it or pay for it.
• Email waitlists. Signing up for a waitlist is one of the lowest-friction actions on the internet. 500 email addresses proves that your landing page headline was interesting. It does not prove that people will transact.
• Friends and family feedback. The people who care about you will support your idea. That is not a signal about the market.
• The "no competitor" argument. Finding no direct competitors is not validation — it might mean the market does not exist, not that you are first.

The three questions that actually matter

Marketplace validation comes down to three questions. You need real evidence for each.

1. Is there real supply?

Can you recruit suppliers who will actually list on your platform, right now, before there is significant demand? Not "would you list if there were buyers" — that is a hypothetical. Would they list today?

The test: manually recruit 20 potential suppliers. Call them. Have a real conversation. Explain that you are building a platform and ask if they will create a profile or listing now. If you cannot get 20 suppliers manually in your first two weeks, you have a supply-side problem that no amount of product will fix.

2. Is there real demand?

Are there buyers actively looking for what your suppliers offer — in a way that is currently underserved? The key word is active. Not "interested in someday" — actively looking, right now, without a good solution.

The test: find where demand currently goes. What do buyers do today when they want this? If the answer is "nothing — there is no solution," ask why. If the answer is "they use [existing solution]," understand why that solution is inadequate. Real demand exists in the gap between what people need and what currently serves them.

3. Will they transact?

This is the question most founders skip — and the one that matters most. Having supply and demand is not enough. They have to actually complete transactions on your platform, at a price point that works for both sides, in a flow they will trust.

The test: facilitate a transaction manually before you build anything. This is the concierge method, and it is the most valuable validation you can do.

The concierge method

The concierge method means doing the work of your marketplace manually before building the technology to automate it. You are the platform.

Find a buyer. Find a seller. Match them. Facilitate the transaction. Handle the payment. Get feedback from both sides.

This sounds slow and unscalable. That is the point. You are not trying to scale yet — you are trying to learn whether the transaction works at all. Does the buyer trust the seller? Does the seller price appropriately? Do both sides find the transaction valuable enough to repeat? What friction exists that you did not anticipate?

Airbnb did this. The founders personally photographed early listings to improve listing quality. They were the platform before the platform existed. The lesson was not that manual processes scale — it was that they learned things from direct involvement that they could never have learned from analytics.

The signals that matter

Once you have done a few manual transactions, here is what you are looking for:

• Repeat behaviour. Did the buyer come back and ask for another transaction? Did the seller ask when they could list more? Repeat behaviour before you have a product is the strongest signal you can get.
• Willingness to pay. Did the buyer pay without negotiating? Did the seller accept the pricing model? Friction around money is a red flag that the value proposition needs work.
• Referral behaviour. Did anyone tell someone else about what you were doing — unprompted? Word of mouth before you have a product means you are solving a real problem for real people.
• The problem they did not tell you about. What came up in the transaction that you did not anticipate? These surprises are the most valuable input you will get in the entire validation process.

How long should validation take?

Four to six weeks, done seriously. You should be able to facilitate 5-10 manual transactions and have real conversations with 20-30 people on each side of the market in that time.

If you cannot get to 5 real transactions in six weeks, that is a signal worth taking seriously before you build.

What good validation looks like

You have validated your marketplace idea when:

• You have facilitated at least 5 real transactions manually
• At least 3 of those buyers said they would use the platform again
• You have 20+ suppliers who have agreed to list when you launch
• You understand the one or two pieces of friction that your platform needs to eliminate to make this work at scale
• You know your liquidity threshold — the minimum supply density at which the buyer experience becomes reliably good

You do not need to have all the answers. You need to have done enough real transactions to know which questions actually matter.

Most founders skip validation because it feels slower than building. In my experience, four weeks of serious validation saves six months of building the wrong thing. It is the best investment you can make before you write a single line of code.