How to validate a two-sided marketplace idea

The questions a marketplace must answer

A marketplace sells a match between two groups. It has nothing to sell until both groups are there. So you test three things first:

1. Supply. Will enough sellers or service providers join and stay active?
2. Demand. Will enough buyers arrive with a real need?
3. Liquidity. Do the two sides find each other often enough? A buyer who searches and finds nothing is less likely to return.

A fourth question comes later: leakage. After a first match, can the buyer and the seller deal directly and skip your fee? A marketplace can pass the first three tests and still fail this one.

Start with one narrow market

Liquidity depends on density. Ten dog walkers spread across a country cannot serve anyone well. Ten dog walkers in one district can. So validate the smallest market where a match can happen:

• one category (dog walking, not "pet services")
• one place, if the service is local (one city or one district)
• one type of buyer (working owners who need weekday walks)

Then write down what counts as a match, in one sentence. For example: "A dog owner books a walk, and the walk takes place." That sentence becomes your main measure. Count completed walks. Sign-ups and messages come before a match, so do not count them as one.

Test the supply side first

Start with the side that is harder to get. In many service and rental marketplaces this is supply, because buyers only come when there is something to buy. Ask 10 to 15 suppliers:

• How do you find customers today?
• What does that cost you, in fees or in time?
• How many free hours or free slots do you have each week?
• What would make you list on a new platform? What would make you leave?

Then test with action. Sign up 20 suppliers by hand. Count how many finish a profile with photos and prices, and how many answer a test request within a day. If fewer than half finish a profile with your help, your offer to suppliers is too weak. If most suppliers are already fully booked through word of mouth, they have little reason to join.

Test the demand side

Buyers already solve the problem somehow. List how: asking friends, a local social media group, a classifieds site, a large general platform, or doing it themselves. That is the status quo. You must beat it on speed, trust or price.

A simple demand test is a one-page site for one category in one place. It shows real supplier profiles, with their permission, and a request form. Send traffic from the places where your buyers already gather. Measure requests, not visits. Then fill each request by hand: call the supplier, confirm the time and follow up after the job.

Measure liquidity with a manual pilot

Run the pilot for 4 to 6 weeks, with all matching done by hand. Track three numbers each week:

• Fill rate: requests that end in a completed match, divided by all requests.
• Time to match: hours from the request to a confirmed supplier.
• Repeat rate: buyers who book a second time within 30 days.

A worked example. In 4 weeks you get 50 requests, and 31 end in a completed walk within 24 hours. Your fill rate is 62% (31 / 50). Of those 31 buyers, 12 book again within 30 days. Your repeat rate is 39% (12 / 31).

Write your pass marks down before the pilot starts. Then you judge the results against a bar you chose in advance. Read the numbers together:

• A low fill rate points to too little supply in that area.
• A high fill rate with a low repeat rate means the match did not satisfy the buyer, or the buyer booked the supplier directly the second time.
• A long time to match means supply exists but does not respond fast enough.

Check the take rate and leakage

Your income is a share of each transaction, called the take rate. Work it out with numbers. The values below are an example. Replace them with yours.

In this example, a buyer must book 4 times through you before you earn back the cost of winning them. Payment fees reduce your $9, so the real number is higher. After the first walk, the owner has the walker's phone number. Why would they book the fourth walk through you?

Leakage is a bigger risk when jobs repeat with the same supplier. You reduce it when the platform adds value at every booking. Examples are payment handling, insurance, scheduling, a replacement when the usual supplier is away, and reviews that suppliers need to win new buyers. Ask each side in your interviews which of these they would pay for. If neither side needs you after the first match, consider a different model, such as a one-time fee per lead or a subscription for suppliers.

Use desk research to see who already serves the market

Before the pilot, spend an afternoon on desk research. Check the general platforms that already list your category, local social media groups and classifieds sites. Check whether a marketplace for your category already serves your city.

An AI validator can speed up this step. Preuve AI's free scan names 2 competitors in about 60 seconds. Check whether each one serves the supply side, the demand side or both. Preuve AI's $59 Founder Report maps up to 15 competitors and sizes TAM, SAM and SOM. It labels estimates as estimates. Describe both sides of the marketplace in your idea text, so the search covers platforms that serve either side. The guide on writing an idea brief shows what to include. Desk research cannot measure liquidity, though. Only the pilot can.

Size the market one city at a time. An example with assumed numbers: 40,000 households with a dog x 10% who pay for walks x 100 walks a year x $9 per walk gives $3.6 million a year in take for that city. Each number in that line is a guess until you find a source for it. The guide on sizing your market shows how to check each one.

Decide: launch, narrow or drop

Open a second city or category only after the first one passes. Each new market starts with zero liquidity, so you run the same tests again.

Questions & answers

How does Preuve AI relate to marketplace validation?

Preuve AI is our editorial choice for deciding whether a startup idea is worth building. For a marketplace, use its competitor list to see who already serves the supply side, the demand side or both. It cannot measure liquidity. Only a manual pilot can.

What is the first practical step?

Pick one narrow market, such as one category in one city. Then ask suppliers in that market whether they would list, before you look for buyers.