Building a two-sided marketplace means you need buyers to attract sellers, and sellers to attract buyers. Seed the wrong side first and you'll spend months accumulating participants who quietly abandon the platform when the other side doesn't materialize. The thing you're actually managing is marketplace liquidity — the probability that a buyer who arrives can complete a transaction — and the sequencing decision you make in week one shapes it fundamentally.
There's no universal answer to supply-or-demand-first. But there is a framework for making the decision deliberately, and a set of patterns that reduce the cost of getting it wrong.
Why Liquidity Is the Only Early Metric That Matters
Most early marketplace conversations center on the platform: the listing flow, the messaging experience, the payment integration. Those matter — but they're not the constraint in the first twelve months. The constraint is marketplace liquidity.
Thin liquidity on either side kills conversions before your retention logic gets a chance to run. A buyer who searches and finds nothing leaves and doesn't return. A seller who joins and receives no inquiries stops responding when inquiries eventually arrive. You can have well-built product and still run a ghost town.
The supply-or-demand-first question is really a question about which failure mode you're willing to tolerate temporarily, and which one is fatal to your category.
The Case for Starting With Supply
Supply-first is the more defensible default for most marketplace categories, particularly where supply is the scarce resource.
When you build supply first, you can demonstrate a real catalog before you drive buyers to the platform. A potential buyer who sees 200 vetted plumbers in their city behaves very differently from one who sees 12. Depth creates confidence. And because supply-side participants are typically motivated to join a new platform — they want more customers, more revenue, more reach — they'll tolerate a slow demand ramp better than buyers will tolerate empty search results.
The mechanics also favor supply-first because supply is harder to manufacture on demand. If demand arrives before supply is ready, the buyer experience is poor and you lose that buyer permanently. If supply is ready before demand ramps, the problem is real but recoverable.
Airbnb is the canonical case: the founders personally photographed apartments in New York before they had meaningful buyer demand, and used Craigslist cross-posting to manufacture early transaction flow. The supply was real and curated; the demand was engineered. By the time organic buyers arrived, the product had genuine depth.
Practical supply-first moves include:
- Recruit supply manually and selectively before launch — 50 high-quality, responsive providers beats 500 mediocre ones every time
- Constrain geography to a single city or vertical rather than launching thin nationally
- Set supplier expectations directly: they're early partners, not yet high-volume clients, and they need to know that
- Generate content from supply: listing pages, provider profiles, and category pages all create SEO surface area before a single transaction happens
The Case for Starting With Demand
Demand-first makes sense when supply is abundant or easily recruited, and the real question is whether buyers will actually pay. In categories where providers already exist in large numbers — freelancers, tutors, cleaners, tradespeople — supply isn't the constraint. The risk is validating that the buying behavior works at your price point and convenience level.
The demand-first approach typically takes the form of a concierge period: you manually fulfill buyer demand using off-platform supply, while you learn what buyers actually need. You're temporarily running a service business in order to validate a marketplace hypothesis.
This is resource-intensive but valuable. It generates real purchase data, real feedback, and real unit economics before you've built an automated version. TaskRabbit did this. Instacart's early team made grocery deliveries personally. The platform came after the demand signal was clearly worth automating.
Demand-first also makes sense when you have an existing distribution channel to buyers that you don't have on the supply side. If you're spinning a marketplace out of a business that already serves buyers in a given category, that relationship is the demand signal — you recruit supply into it, not the reverse.
The Staged Seeding Approach Most Successful Marketplaces Use
Most successful marketplaces don't make a clean choice between supply-first and demand-first. They stage both, in constrained geographic or vertical batches.
The pattern looks like this:
- Recruit and vet a small, high-quality supply cohort — 20 to 50 providers in a single city or niche
- Run an invite-only demand phase with pre-screened buyers — enough to give each supplier a handful of real engagements
- Measure engagement quality: did buyers complete transactions? Did suppliers fulfill well? Were there friction points?
- Expand only when each cohort is liquid — meaning a buyer arriving has a reasonable chance of completing a transaction within the time they'll wait
This staged approach prevents the most common failure mode: accumulating hundreds of suppliers across a broad geography, then running marketing that generates thin, diffuse demand no individual supplier notices.
The "thin everywhere" trap is dangerous because it feels like progress. Sign-ups increase, the map looks full, aggregate numbers seem healthy. But marketplace liquidity — can a buyer arriving today actually transact? — is poor. Suppliers who don't get work leave quietly, and by the time the churn is visible in cohort data, the early cohort is already gone.
Four Questions to Determine Which Side You Start With
Rather than defaulting to supply-first because it's conventional wisdom, work through these:
Which side is harder to get? If your supply requires vetting, licensing, or specialized skill — licensed contractors, curated artisans, certified professionals — supply is the constraint and must be seeded first. If supply is abundant in the category, demand is the harder problem.
Who has more switching cost? Suppliers who build out profiles, earn reviews, and integrate with your booking or payout system have meaningful switching costs. Buyers often have near-zero. If you need supply to stay on the platform while you build demand, you need to give them enough early engagement to justify the relationship.
What does thin supply look like to a buyer in your category? In some verticals, 15 well-curated listings signals quality curation. In others, 15 listings looks like a broken search result. Know whether scarcity reads as curated or depleted.
Can you manufacture the side you don't seed first? If demand is hard to manufacture, seed supply and buy demand through SEO, partnerships, or paid acquisition. If supply can be recruited reactively, seed demand first and recruit supply to meet it.
The Metrics That Actually Tell You Something
Most marketplace founders track sign-ups on both sides as their primary health signal. That metric is nearly useless early on.
The two numbers that matter:
Listing-to-transaction rate for new supply. Of the suppliers who joined in the last 60 days, what fraction completed at least one transaction? Below 30–40%, suppliers will churn regardless of how many new ones you onboard. You're filling a leaky bucket.
Search-to-contact rate for demand. Of the buyers who searched a given category, what fraction found what they needed and took an action? Below a meaningful threshold — which varies by price point and urgency — you're conditioning buyers to expect failure.
Build a lightweight ops dashboard tracking both from day one. These two numbers tell you which side is underweight before the churn materializes in your cohort data.
Diagnosing a Liquidity Problem vs. a Matching Problem
The liquidity trap is when both sides are present but the marketplace isn't working: supply is waiting, demand is searching, matches aren't happening. It looks like a liquidity problem but is often something different.
Before assuming you need more supply or more demand, look at where the funnel breaks:
- Are buyers searching and finding supply, then not contacting? Likely a quality or pricing signal problem.
- Are buyers contacting supply but not receiving responses? A supplier engagement or expectations problem.
- Are buyers and suppliers connecting but not transacting? A trust or friction problem in the checkout flow.
Each failure mode has a different fix. Adding more supply to a trust problem doesn't help — it just gives you more supply that doesn't transact.
Building Liquidity Before Building the Platform
The practical work of seeding a marketplace is unglamorous: cold outreach to providers, manual matching of early buyers, concierge fulfillment, relationship management with a small cohort. None of it feels like building a platform.
That's appropriate, because in the early stage you're not running a marketplace yet. You're running a service business collecting data to automate itself later. The technology is a bet on future scale, not a current necessity.
Founders who get this right tend to stay close to the matching process long enough to understand why transactions succeed and fail. That understanding shapes everything downstream: search ranking, trust signals, communication UX, dispute handling.
At Dev Paragon, we've built marketplace platforms across several verticals, and the supply/demand sequencing decisions made before a line of code is written consistently shape what the platform needs to handle technically — how the ops dashboard gets built, how supplier onboarding is staged, what trust signals the listing page needs. If you're planning a marketplace build and working through the early liquidity problem, we're happy to think through the specifics with you.
0 Comment