You want to build the next Airbnb. Uber. Etsy. A platform that connects two sides: drivers and riders, hosts and guests, sellers and buyers. The rewards are enormous. The challenge is even bigger.
The chicken-and-egg problem: you need supply to attract demand. You need demand to attract supply. You have neither. You start with nothing.
Most marketplace founders fail because they try to launch both sides at once. They build a beautiful platform. They open the doors. Nobody comes. The supply side says “not enough demand.” The demand side says “not enough supply.” Everyone waits. Everyone loses.
The winners do something different. They fake one side. They subsidize the other. They launch in a tiny geography. They do things that do not scale. They solve the chicken-and-egg problem with creativity, not capital.
Here is how to launch a multi-sided marketplace. From zero to liquidity. From nothing to network effects.
The Story That Proves the Point
Let me tell you about the ride-sharing company that launched with no drivers and no riders.
Uber did not launch in 100 cities. They launched in one city: San Francisco. They did not hope drivers would show up. They recruited drivers manually. They went to car washes, taxi stands, and limo companies. They begged drivers to sign up.
They did not hope riders would show up. They stood outside tech conferences and gave free rides. They went to bars at 2 AM and offered to drive people home. They did things that did not scale.
For months, Uber was a fake marketplace. They subsidized both sides. They paid drivers even when there were no riders. They gave free rides even when it cost them money.
Then something happened. Enough drivers. Enough riders. The marketplace hit liquidity. Riders found drivers instantly. Drivers found riders instantly. The network effects kicked in.
Uber did not solve the chicken-and-egg problem with a brilliant algorithm. They solved it with manual labor, subsidies, and a single city.
Here is how you can do the same.
The Chicken-and-Egg Problem (And Why It Is So Hard)
A multi-sided marketplace has two (or more) customer groups. Each group needs the other to derive value.
Examples:
-
Airbnb: Hosts need guests. Guests need hosts.
-
Uber: Drivers need riders. Riders need drivers.
-
Etsy: Sellers need buyers. Buyers need sellers.
-
DoorDash: Restaurants need diners. Diners need restaurants.
The problem: each group will not join until the other group is already there. This is the chicken-and-egg problem.
Why it is hard:
-
You cannot build both sides at once (too expensive, too slow).
-
You cannot launch to “everyone” (too diffuse, no liquidity).
-
You cannot wait for organic growth (it will not come).
The solution: Launch in a way that does not require both sides to be perfect. Falsify one side. Subsidize the other. Launch tiny. Do things that do not scale.
Step 1: Choose a Single, Tiny Geography
The biggest mistake marketplace founders make is launching too broadly. “We will launch in the US.” No. Launch in one neighborhood. One city block. One college campus.
Why tiny geography works:
-
Density creates liquidity. Ten drivers and ten riders in one neighborhood find each other instantly. Ten drivers and ten riders spread across a city do not.
-
Marketing is cheaper. You can knock on doors. You can hang flyers. You can talk to humans.
-
You can subsidize affordably. Free rides for 100 people costs less than free rides for 10,000.
Examples:
-
Uber launched in San Francisco (not California, not the US).
-
Airbnb launched in San Francisco during a design conference (one neighborhood, one weekend).
-
DoorDash launched on the Stanford University campus (one school, one zip code).
Your job: Pick one neighborhood. One zip code. One apartment building. Master it. Then expand.
Step 2: Solve One Side First (Usually Supply)
You cannot solve both sides at once. Pick one side to solve first. Usually, solve supply first.
Why supply first:
-
Demand comes when supply is ready. A rider who opens your app and sees no drivers will never return. A guest who sees no listings will never return.
-
Supply is harder to recruit. Drivers need training. Hosts need listings. Sellers need inventory. Do the hard thing first.
How to solve supply first:
-
Recruit manually. Knock on doors. Make phone calls. Offer incentives.
-
Subsidize supply. Pay drivers even with no riders. Give hosts free photography.
-
Do things that do not scale. Help sellers list their first product. Write their descriptions for them.
Example (DoorDash): Before launching, DoorDash founders went to restaurants and asked for their menus. They typed them up manually. They took photos of every dish. They convinced restaurants to participate by promising free delivery. They solved supply first.
Step 3: Falsify the Other Side (Pretend Demand Exists)
Once you have supply, you need demand. But demand will not come until supply is proven. So falsify demand.
What “falsify demand” means:
-
Pretend you have more demand than you actually do.
-
Subsidize the demand side to create activity.
-
Do things that look like demand until real demand arrives.
How to falsify demand:
-
Offer free or discounted service to early demand-side users.
-
Run promotions that make it look like others are using the platform.
-
Create a waitlist that signals scarcity and popularity.
Example (Uber): Uber gave free rides at tech conferences. They created lines of people waiting for rides. Real demand was low. Perceived demand was high. Drivers saw lines and thought “this is busy.” More drivers signed up. More drivers meant shorter waits. Shorter waits brought real demand.
Pro tip: Do not lie. Falsifying demand means creating real activity through subsidies, not faking data. Give free rides. That is real demand (just subsidized).
Step 4: Do Things That Do Not Scale
Marketplaces require manual effort in the beginning. Automate later. Do not automate too early.
Things that do not scale:
-
Recruiting supply one by one (calls, emails, in-person visits).
-
Onboarding supply manually (help them list, photograph, describe).
-
Fulfilling demand yourself (be the first driver, host, or seller).
-
Collecting feedback in person (ask every user what they think).
-
Fixing problems immediately (not with a ticket system, with a text message).
Example (Airbnb): The founders flew to New York. They rented a camera. They visited every host’s apartment. They took professional photos. They wrote better descriptions. They did this for months. It did not scale. It worked.
Pro tip: Ask yourself: “What would I do if I had only 100 users?” Do that. Do not build features. Do not automate. Do human things.
Step 5: Subsidize One Side (Usually Demand)
You will need to spend money to solve the chicken-and-egg problem. Plan for it. Budget for it.
Why subsidize demand:
-
Demand is price-sensitive. Lower prices bring more users.
-
Demand creates activity. Activity attracts supply.
-
Demand-side subsidies are easier to measure (cost per acquisition, cost per transaction).
How to subsidize demand:
-
Free trials (first ride free, first night free, first purchase free).
-
Discount codes (50% off first three rides).
-
Referral credits (give $10, get $10).
-
Loss-leading pricing (charge below cost temporarily).
How to subsidize supply (if needed):
-
Guaranteed minimum earnings (“earn at least $20/hour even with no rides”).
-
Free equipment (delivery bags, signage, photography).
-
Lower commissions (0% for first 100 sales).
Pro tip: Subsidize the side that is harder to acquire. For most marketplaces, demand is harder. For some (like high-end service marketplaces), supply is harder. Know your business.
Step 6: Launch with a “Seeded” Marketplace
Do not launch with zero of anything. Seed the marketplace before you open the doors.
How to seed supply:
-
Recruit 10–20 supply-side users before launch.
-
Pay them to be ready (even with no demand).
-
Create fake demand if needed (internal test orders).
How to seed demand:
-
Recruit 50–100 demand-side users before launch.
-
Give them free or discounted service.
-
Create a waitlist that signals popularity.
The seeding number: For a local marketplace (rides, food delivery, services), aim for 10–20 supply and 50–100 demand before you “launch.” For a national marketplace, start with one city and hit those numbers there.
Pro tip: Do not announce your launch date publicly. Launch quietly. Fix problems with real users. Then announce when things work.
Step 7: Measure Liquidity, Not Users
Users do not matter. Liquidity matters. Liquidity is the ability for a demand-side user to find a supply-side user quickly and reliably.
Liquidity metrics:
-
Match rate: What percentage of demand-side requests get fulfilled?
-
Time to match: How long between request and fulfillment?
-
Fill rate: What percentage of supply-side inventory gets used?
Example (Uber):
-
Bad: 10,000 drivers and 10,000 riders spread across a city (low liquidity).
-
Good: 100 drivers and 100 riders in one neighborhood (high liquidity).
What to measure:
-
For a rides marketplace: “What percentage of ride requests get a driver within 5 minutes?”
-
For a rental marketplace: “What percentage of searches show at least 3 available listings?”
-
For a job marketplace: “What percentage of job posts get at least 5 applications?”
Pro tip: Do not add a new city until your first city has high liquidity. Premature expansion kills marketplaces.
The Marketplace Launch Checklist
Before launch (2–3 months):
-
Choose one tiny geography (neighborhood, campus, zip code).
-
Recruit 10–20 supply-side users manually.
-
Seed the supply side with subsidies (pay them to be ready).
-
Build a simple platform (MVP, not perfect).
-
Recruit 50–100 demand-side users (friends, family, early testers).
-
Run manual tests (you be the driver, host, or seller).
Launch day:
-
Do not announce widely. Launch quietly.
-
Fulfill every request manually if needed.
-
Fix problems in real time (text message, phone call).
-
Collect feedback from every user.
First 30 days after launch:
-
Subsidize demand (free rides, discounts, credits).
-
Recruit more supply based on demand signals.
-
Measure liquidity daily (match rate, time to match).
-
Do not add features. Fix what is broken.
-
Do not expand geography. Master your first market.
After liquidity is achieved:
-
Reduce subsidies slowly (test price sensitivity).
-
Add automation for manual processes.
-
Expand to adjacent geography (next neighborhood).
-
Repeat the process.
A Real-World Example: The Local Services Marketplace
A founder named Sarah wanted to build a marketplace connecting homeowners with plumbers, electricians, and handymen.
Step 1 (Tiny geography): She chose one zip code. 10,000 homes. 50 service providers.
Step 2 (Solve supply first): She visited every plumber, electrician, and handyman in that zip code. She offered free listings for one year. She helped them write descriptions and upload photos. She recruited 15 providers.
Step 3 (Falsify demand): She posted flyers in coffee shops and laundromats. “Need a plumber? Get $20 off your first booking.” She gave her first 10 customers $50 credits.
Step 4 (Do things that do not scale): When a homeowner requested a plumber, Sarah called the plumber herself. She confirmed availability. She followed up after the job. She asked for feedback.
Step 5 (Subsidize demand): She charged homeowners nothing for the first 50 bookings. She paid plumbers full price out of her own pocket. She lost money on every transaction.
Step 6 (Seeded marketplace): By week 4, she had 20 providers and 100 homeowners. Match rate was 80%. Time to match was under 2 hours.
Step 7 (Measure liquidity): She tracked “percentage of requests fulfilled within 2 hours.” When it hit 90%, she knew she had liquidity.
Sarah then expanded to the next zip code. Then the next. Within 18 months, she had 5,000 providers and 50,000 homeowners.
She did not launch nationally. She launched in one zip code. She did things that did not scale. She solved the chicken-and-egg problem.
Frequently Asked Questions (FAQ)
How much money do I need to launch a marketplace?
Less than you think. Uber launched with a few hundred thousand dollars. Airbnb launched with credit card debt. Start with enough to subsidize one tiny geography. $10,000–$50,000 is often enough for an MVP.
What if I have no money to subsidize?
Barter. Give free listings to supply in exchange for exclusivity. Give free service to demand in exchange for referrals. Trade equity for early supply. Get creative.
How do I know if my marketplace has product-market fit?
When demand-side users come back without subsidies. When supply-side users stay without guarantees. When liquidity happens naturally. When you can reduce subsidies and usage does not drop.
Should I build a mobile app first?
No. Build a website first (or even a Google Sheet + text messages). Uber started with a website and BlackBerry app. Airbnb started with a website. Build the simplest possible thing. Add complexity later.
What is the biggest mistake marketplace founders make?
Expanding too fast. Adding too many features. Automating too early. Launching too broadly. Not doing things that do not scale. Master one tiny geography before you do anything else.
The Bottom Line
Multi-sided marketplaces are the hardest businesses to start. They are also the most valuable.
The chicken-and-egg problem is real. But it is solvable.
-
Choose one tiny geography (neighborhood, campus, zip code).
-
Solve supply first (recruit manually, subsidize, do not scale).
-
Falsify demand (free rides, discounts, waitlists).
-
Do things that do not scale (manual onboarding, personal calls).
-
Subsidize the harder side (usually demand).
-
Seed the marketplace before launch (10 supply, 50 demand).
-
Measure liquidity, not users (match rate, time to match).
Do not launch in 100 cities. Launch in one block. Do not build for millions. Build for dozens. Do not automate. Do human things.
Liquidity is not magic. It is density. Density comes from focus.
Master one tiny market. Then expand. Then conquer.
Ready to launch your multi-sided marketplace? Share this post with a fellow founder who needs to see it. And subscribe to our newsletter for more marketplace strategies every Tuesday.
