Every two sided marketplace faces the same fundamental problem: buyers will not come without sellers, and sellers will not come without buyers. This is not a hypothetical business school exercise. It is the single biggest reason marketplace startups fail. We have built marketplaces that solved this problem and worked with teams whose marketplaces did not survive it. The difference is almost never the technology. It is the launch strategy.
We have written about the technical architecture of marketplace apps separately. This post focuses on the strategic and tactical approaches to the cold start problem, along with the technical infrastructure each approach requires.
Why Marketplaces Are Worth the Pain
Before diving into solutions, it is worth understanding why marketplaces are so valuable when they work. A successful two sided marketplace creates network effects: each new seller makes the platform more valuable to buyers, and each new buyer makes it more valuable to sellers. Once you reach critical mass, this flywheel becomes a moat that is nearly impossible for competitors to replicate.
Uber, Airbnb, and Etsy did not win because they had better technology. They won because they solved the chicken and egg problem first in their respective markets. The technology needs to be good enough, but the launch strategy is what determines whether you reach the network effect tipping point.
Strategy 1: Start With Supply
The approach: Recruit sellers first, then attract buyers to an existing catalog.
This is the most common approach and it works well when your supply side has independent value. Etsy recruited craft sellers who already had products to sell. Airbnb initially targeted people who were already listing on Craigslist. The supply existed; the marketplace just gave it a better home.
How to execute:
- Manually recruit your first 50 to 100 sellers. This does not scale, and that is fine. In the early days, the founders should be personally onboarding every seller.
- Offer zero fees or negative fees (subsidized listings) for the first cohort. Your first sellers are the most valuable users you will ever have.
- Pre populate their listings. If sellers need to upload photos, write descriptions, and set prices, the onboarding friction will kill your supply growth. Do the work for them initially.
Technical requirements:
- A seller onboarding flow that is as frictionless as possible. We build these with progressive profiling: collect the minimum information upfront and enrich later.
- Bulk import tools so you can onboard sellers from spreadsheets or existing platforms.
- Managed listings where your team can create and edit listings on behalf of sellers.
- A CMS layer that lets you curate and present sparse inventory in a way that does not look empty.
Strategy 2: Start With Demand
The approach: Build an audience of potential buyers first, then bring on sellers to serve them.
This works when you can provide value to the buy side before the marketplace exists. Thumbtack started by offering a free service request form. Customers submitted what they needed, and Thumbtack manually matched them with service providers. The marketplace evolved from a lead generation tool.
How to execute:
- Build a content or tool offering that attracts your target buyers. This could be a blog, a comparison tool, a free resource, or a community.
- Collect demand signals: wishlists, service requests, pre orders, or intent data.
- Use the demand data to recruit sellers: "We have 500 people looking for X in your area. Want to be listed?"
Technical requirements:
- A demand capture system (waitlists, request forms, wishlists) that stores structured intent data.
- Analytics and reporting that lets you show sellers exactly how much demand exists for their category or location.
- An email notification system that alerts early registrants when matching supply becomes available. We use Supabase edge functions for this kind of event driven notification pipeline.
- A landing page framework that can be customized by category or geography to capture targeted demand.
Strategy 3: Single Player Mode
The approach: Build a product that is valuable to one side of the marketplace even without the other side.
This is the most elegant solution because it removes the chicken and egg problem entirely. OpenTable started as a reservation management tool for restaurants (supply side value) before adding the consumer facing booking feature. Shopify started as an e commerce tool for individual sellers before building a marketplace.
How to execute:
- Identify a pain point on one side of the marketplace that your platform can solve independently.
- Build a SaaS tool for that pain point and acquire users.
- Once you have a critical mass of users on one side, layer in the marketplace functionality.
Technical requirements:
- A standalone product that works without the marketplace component. This is a full product build, not a marketplace build.
- Data infrastructure designed so that user activity in single player mode generates the data needed for marketplace matching later.
- A gradual feature rollout system (feature flags) that lets you introduce marketplace functionality to subsets of users.
- Integration architecture that allows the SaaS tool and marketplace to share data and user identity seamlessly.
This approach requires more upfront development but de risks the marketplace launch. We cover the broader product building process in our guide to building an MVP.
Strategy 4: Geographic or Category Constraint
The approach: Launch in a tiny market where you can achieve liquidity quickly, then expand.
Uber launched in San Francisco only. Airbnb focused on conference cities during major events. By constraining the market, you need fewer participants on each side to create a functioning marketplace.
How to execute:
- Pick one city, one neighborhood, or one product category.
- Achieve critical mass in that micro market. For most marketplaces, this means at least 30% of potential supply in the area is listed, and buyers can find what they want at least 80% of the time.
- Build the playbook for launching a new market, then repeat.
Technical requirements:
- Geo fencing and location based search that constrains the user experience to the active market.
- A market launch toolkit: automated seller outreach templates, localized landing pages, market specific analytics dashboards.
- Multi tenancy architecture where each market can have its own configuration (pricing, categories, rules) while sharing the core platform.
- Supply density scoring that tells you when a market is ready to launch and where gaps exist.
Strategy 5: Subsidize One Side
The approach: Pay one side of the marketplace to participate until the other side generates enough value.
ClassPass subsidized studios with guaranteed payments. Uber gave drivers minimum earning guarantees. DoorDash paid restaurants upfront fees to join.
How to execute:
- Calculate the maximum subsidy you can afford per unit of supply or demand. This is a runway calculation, not a profitability calculation.
- Structure the subsidy to decrease over time as organic activity grows. Week 1 you pay 100% of the seller's ask. Week 8 you pay 50%. By week 16, the marketplace fee structure applies normally.
- Track the metrics that tell you when a market can sustain itself without subsidies.
Technical requirements:
- A subsidy management system that tracks guarantees, calculates payouts, and tapers subsidy rates on a schedule.
- Financial reconciliation infrastructure that handles the difference between what buyers pay and what sellers receive during the subsidy period.
- Dashboard and reporting for monitoring burn rate per market and per cohort.
- Payment infrastructure that supports split payments, holds, and multi party disbursement. Our payment processing guide covers the technical details.
The Metrics That Tell You It Is Working
Regardless of which strategy you use, these are the numbers we track to determine marketplace health:
- Liquidity rate: What percentage of listings result in a transaction? Below 10% and your marketplace feels dead. Above 30% and you have meaningful liquidity.
- Time to first transaction: How long does it take a new seller to make their first sale, or a new buyer to complete their first purchase? If this exceeds 2 weeks, you have a retention problem.
- Repeat rate: What percentage of buyers make a second purchase within 30 days? This is the strongest signal that your marketplace is providing genuine value.
- Supply side retention: What percentage of sellers are still active after 90 days? If sellers are churning, it does not matter how many buyers you attract.
- Search to fill rate: When a buyer searches for something, how often do they find a relevant result? Below 70% and the experience is frustrating.
Building for the Transition
The biggest technical mistake we see in marketplace development is building the full marketplace infrastructure before proving demand. You do not need a real time matching algorithm, a reputation system, and a dispute resolution workflow before you have 100 users on each side. Start with manual processes, a spreadsheet, and direct communication. Automate once you understand the patterns.
What you do need from day one is a data model that supports marketplace dynamics: separate user types (or a flexible role system), transaction records that link buyer and seller, and event tracking that captures the signals you need to measure the metrics above.
We have built marketplace platforms including GameLootBoxes, where solving the supply and demand balance was central to the product's success. The comparison between custom development and off the shelf SaaS is particularly relevant for marketplace founders evaluating build versus buy decisions for their platform.
If you are building a marketplace and want to pressure test your launch strategy, or if you need the technical architecture to support it, we would like to hear about it.