If you are running a marketplace, you are not running one business. You are running two. You have a supply-side business (sellers, drivers, hosts) and a demand-side business (buyers, riders, guests). And then, you have a third business: the engine that connects them.
“Marketplace analytics” is the art and science of measuring this complex ecosystem. Unlike a simple SaaS company that cares about MRR (Monthly Recurring Revenue) and Churn, or an e-commerce store that cares about conversion rates, a marketplace dies or thrives based on the efficiency of its matches.
Most founders get this wrong. They look at “Total Registered Users” and pat themselves on the back. They track “Total Revenue” and think they are growing. But underneath, their marketplace might be burning.
In this complete guide, we will move beyond vanity metrics. We will explore the specific data frameworks, KPIs, and tools you need to build a liquidity machine.
The Two-Sided Dilemma: Why Standard Analytics Fail
Standard analytics tools like Google Analytics are built for single-sided funnels. A user comes in, they view a product, they buy it.
In a marketplace, the funnel is a loop.
- Supply joins.
- Supply creates inventory.
- Demand joins.
- Demand searches.
- Demand finds Supply.
- Transaction happens.
- Review is left.
- Loop repeats.
If you try to jam this into a standard e-commerce dashboard, you miss the most critical signal: Liquidity. You might have 10,000 buyers and 1,000 sellers, but if they never find each other, your marketplace is effectively zero.
The Hierarchy of Marketplace Metrics
Not all metrics are created equal. Depending on your stage, you should be obsessed with different numbers. We can view this as a hierarchy of needs.
Level 1: Liquidity (The Foundation)
Before you care about revenue, you must care about matching.
- Search to Fill: What percentage of searches result in a transaction?
- Time to Fill: How long does a listing sit before it is booked?
- Zero Search Results: How often do buyers search and find nothing?
If you have low liquidity, you do not have a product yet.
Level 2: Retention (The Validation)
Once matches are happening, do they happen again?
- User Retention: Do buyers come back?
- Provider Retention: Do sellers keep listing?
- Wallet Share: Are you capturing more of their spend over time?
Level 3: Unit Economics (The Sustainability)
Can you make money?
- LTV (Lifetime Value): How much is a user worth?
- CAC (Customer Acquisition Cost): How much does it cost to get them?
- Contribution Margin: Profit per transaction after variable costs.
Level 4: GMV (The Scale)
Only after the first three are healthy should you obsess over Gross Merchandise Value. Growing GMV on bad unit economics is just scaling your losses.
Deep Dive: Measuring Liquidity
Liquidity is the probability of a transaction happening. It is the single most important concept in marketplace analytics.
Buyer Liquidity
“If I search for X, how likely am I to find it?” To track this, you need to instrument your search bar. Every search query is a data point.
- Search Conversion Rate: (Sessions with a Transaction / Total Sessions with a Search).
- Inventory Density: The average number of results returned per search. If this is too low (< 3), buyers perceive a lack of choice. If it is too high, they suffer from decision paralysis.
Seller Liquidity
“If I list Y, how likely am I to sell it?” This dictates your supply churn. If a seller lists an item on eBay and it doesn’t sell in two weeks, they might not list again.
- Utilization Rate: (Hours Booked / Hours Available). Critical for service marketplaces like Upwork or Uber.
- Sell-Through Rate: (Items Sold / Items Listed). Critical for product marketplaces like Poshmark.
Pro Tip: Measure liquidity by cohort and geography. Your marketplace might be liquid in NYC but illiquid in SF. An aggregate number will hide this truth.
The Truth Teller: Cohort Analysis
Aggregate charts always go up and to the right if you are spending money on marketing. They are vanity metrics. To see the truth, you must use Cohort Analysis.
A cohort is a group of users who started at the same time.
The “Smile” Retention Curve
In SaaS, retention usually decays over time (100% -> 80% -> 60% -> 50%). In healthy marketplaces, you often see a “smile.”
- Month 1: 100% (Acquisition)
- Month 3: 40% (Drop off)
- Month 12: 50% (Reactivation)
Why? Because marketplaces often have natural frequency. I might only need a plumber once a year. Seeing 0% activity in Month 2-11 doesn’t mean I churned; it means I haven’t had a plumbing issue.
You need to track “Revenue Retention” per cohort.
- Healthy: The 2023 cohort spends more in 2024 than they did in 2023. This is “Negative Net Churn.”
- Unhealthy: The 2023 cohort spends 10% of what they did in 2024. You are constantly replacing old users with new ones.
Supply vs. Demand Funnels
You need two separate dashboards.
The Supply Funnel
- Visitor: Lands on “Become a Seller” page.
- Sign Up: Creates an account.
- Onboarding: Uploads ID, connects bank account.
- Activation: Publishes first listing.
- Transaction: Completes first sale.
Where do they drop off?
- If drop-off is high at Onboarding, your form is too long or your KYC (Know Your Customer) process is broken.
- If drop-off is high at Activation, your listing tool is too complex.
The Demand Funnel
- Visitor: Lands on homepage.
- Searcher: Executes a query.
- Viewer: Clicks a listing.
- Initiator: Clicks “Book” or “Buy”.
- Transactor: Completes payment.
Where do they drop off?
- High drop-off from Search to View? Your search results are irrelevant or your thumbnails are bad.
- High drop-off from View to Initiator? Your prices are too high or descriptions are poor.
- High drop-off from Initiator to Transactor? Your checkout flow is broken or trust is low (hidden fees).
Unit Economics: The LTV/CAC Ratio
Marketplace analytics eventually boils down to: Can you buy money for less than it is worth?
Calculating CAC (Customer Acquisition Cost)
This is tricky in marketplaces because of “Cross-Side Network Effects.” Sometimes you spend money to acquire Supply, and that Supply brings their own Demand (e.g., a yoga teacher brings their students to a platform).
Blended CAC = Total Marketing Spend / Total New Paying Users
However, you should split this:
- Supply CAC: Cost to get a seller to list.
- Demand CAC: Cost to get a buyer to buy.
Calculating LTV (Lifetime Value)
LTV = Average Order Value (AOV) * Purchase Frequency * Commission Rate * Lifetime
Example:
- AOV: $100
- Frequency: 4 times/year
- Commission: 10% ($10 net revenue)
- Lifetime: 3 years
- LTV: $10 * 4 * 3 = $120
If your CAC is $50, your LTV/CAC ratio is 2.4x. This is okay, but not great. Investors usually look for 3x or higher.
The Modern Data Stack for Marketplaces
How do you actually track all this?
Stage 1: The MVP (0 - 1,000 Users)
Do not over-engineer.
- Database: Your production database (Postgres/MySQL).
- Visualization: Metabase or just SQL queries.
- Tracking: Google Analytics 4 (for basic traffic).
Stage 2: The Growth Phase (1,000 - 100,000 Users)
You need event tracking.
- Collection: Segment (to capture events like “Listing Created”, “Booking Request”).
- Product Analytics: Mixpanel or Amplitude. These tools are great for funnel analysis and retention cohorts.
- Transactional Email: Customer.io or Braze (triggered by data).
Stage 3: Scale (100,000+ Users)
You need a Data Warehouse.
- Warehouse: Snowflake or BigQuery.
- ETL: Fivetran (to move data from Stripe/Postgres to Warehouse).
- BI Tool: Looker, Tableau, or dedicated marketplace tools like Twosided.
Common Analytics Mistakes
1. Focusing on Registered Users
A “user” who hasn’t transacted is a cost, not an asset. Focus on “Active Buyers” and “Active Sellers.”
2. Ignoring Seasonality
Marketplaces are often seasonal. Airbnb peaks in summer. Uber peaks on New Year’s Eve. Compare Year-Over-Year (YoY), not Month-Over-Month (MoM), to account for this.
3. Blended Averages
“Average Order Value is $50.” This is dangerous. You might have 90% of orders at $10 and 10% at $500. The “average” user doesn’t exist. Always segment your data (e.g., “High Value” vs “Low Value” users).
Conclusion
Marketplace analytics is not just about generating reports; it is about finding the constraints in your network.
Is your constraint Supply? (High search failure, low inventory). Is your constraint Demand? (High inventory, low utilization).
Data tells you where to focus. If you are Supply constrained, stop spending money on Facebook Ads for buyers. Go recruit sellers.
By rigorously tracking Liquidity, Retention, and Unit Economics, you can turn a chaotic bazaar into a predictable, scalable machine. The companies that win are not always the ones with the best idea, but the ones with the clearest view of their own ecosystem.
Stop guessing. Twosided is the analytics platform built specifically for marketplaces. We integrate with your database to give you instant access to Liquidity, Cohort, and GMV dashboards. Get your free demo.