The 13 Metrics That Actually Reveal If Your Marketplace Is Working — And What Each One Signals
Andreessen Horowitz partners Jeff Jordan, Li Jin, D'Arcy Coolican, and Andrew Chen published a framework of 13 metrics specific to marketplace businesses. The piece distinguishes marketplace measurement from general startup metrics, arguing that standard SaaS or e-commerce benchm
What Happened
Andreessen Horowitz partners Jeff Jordan, Li Jin, D'Arcy Coolican, and Andrew Chen published a framework of 13 metrics specific to marketplace businesses. The piece distinguishes marketplace measurement from general startup metrics, arguing that standard SaaS or e-commerce benchmarks fail to capture what actually matters in two-sided platforms. It covers everything from match rates and market depth to multi-tenanting, switching costs, and power user curves. The framework is designed to help founders both diagnose current health and evaluate long-term defensibility.
Why It Matters
Most marketplace founders default to vanity metrics — GMV, total users, app downloads — that obscure whether the core matching engine is actually working. The deeper signal here is that marketplace health is not one number; it is a system of interconnected signals. A high GMV with a low match rate means you have traffic but no liquidity. Strong retention in old cohorts but declining retention in new ones means your early users were outliers, not proof of product-market fit. What this framework reveals is that marketplaces can appear to be growing while silently deteriorating — and founders who only watch GMV will miss it until it is too late. Understanding how to build a successful marketplace means learning to track the signals that actually reflect whether your matching engine is working, not just the numbers that look good in a pitch deck.
Marketplace Insight
SUPPLY: Market depth and concentration metrics tell you whether your supply side is deep enough to satisfy demand, and whether you are dangerously dependent on a small number of top suppliers. If your top 10 sellers account for 60% of GMV, one defection breaks your marketplace. For heterogeneous supply (where every listing is unique, like Airbnb), more supply keeps adding value. For homogeneous supply (like scooters), past a certain density threshold, more supply adds no incremental value to users.
DEMAND: Match rate is the single most honest signal of demand-side satisfaction. Tracking 'zeros' — users who opened the app or searched but did not transact — tells you where demand is leaking. High zero rates point to supply gaps, pricing friction, or trust failures, not just a discovery problem.
LIQUIDITY: Time to match (days to turn) is the operational heartbeat of liquidity. A marketplace where supply sits unsold for weeks, or where buyers cannot find a match quickly, is a low-liquidity marketplace regardless of how much inventory exists. Liquidity is not about volume — it is about speed and clearance rate.
TRUST: Switching costs and multi-homing costs are proxies for trust and lock-in. High switching costs (like Stitch Fix's preference calibration process) are a trust and data moat. Low switching costs (like job boards, where employers post everywhere) signal a commodity marketplace with thin defensibility and constant margin pressure.
GROWTH: User retention cohorts and core action retention cohorts are the real growth indicators. If newer cohorts are retained at higher rates than older ones, your network effects are real and compounding. If retention is flat or declining across cohorts, growth is masking structural weakness — you are filling a leaky bucket.
ONBOARDING: The cold start value question — how much value does a new user get on day one? — directly determines whether your marketplace can grow organically or requires heavy acquisition spend. A marketplace where new users get immediate value (even with a thin network) has a structural onboarding advantage over one that requires network density before delivering value, which is why following marketplace launch best practices from the outset can significantly reduce early churn.
MONETIZATION: Take rate is not just a revenue metric — it is a signal of marketplace power. A high take rate that suppliers accept without leaving indicates real value creation. A take rate being squeezed over time, or one that must be kept artificially low to prevent supplier defection, signals weak positioning. Dollar retention cohorts reveal whether paying users find increasing value over time, which is the clearest monetization health check available.
What This Means for Marketplace Founders
Non-technical founders often lack dashboards or engineers to build custom analytics, which creates a false sense that these metrics are out of reach. They are not. Most of these can be tracked with basic spreadsheets, simple survey questions to users, or tools like Airtable, Mixpanel, or even manual review of transaction logs.
The more important implication is prioritization. You cannot optimize 13 metrics simultaneously. The right approach is to identify which two or three metrics are most broken relative to your stage. Early-stage marketplaces almost always have a liquidity problem first — match rate and time to match should dominate attention before you worry about power user curves or dollar retention cohorts.
Founders also need to resist the temptation to aggregate metrics across markets. If your marketplace operates in multiple cities or categories, a healthy average can hide a failing local market. Track metrics by geography or category segment, especially if your network effects are local. Understanding community marketplace essentials can help founders think more clearly about how local trust and engagement influence these segmented metrics.
Finally, multi-tenanting prevalence is the metric most founders ignore and most regret later. If your users are actively using two or three competing platforms simultaneously, your retention numbers will look acceptable while your actual share of wallet and share of transactions quietly erodes.
Actionable Takeaways
• Define your match rate before anything else. What does a successful transaction look like in your marketplace? Measure the percentage of sessions or searches that result in one. Track your 'zeros' separately and investigate the top three reasons they occur.
• Audit your supply concentration monthly. Calculate what percentage of your GMV comes from your top 10 suppliers. If it exceeds 40–50%, you have a concentration risk that needs to be addressed through supply acquisition before it becomes a leverage problem.
• Measure time to match for both sides. How long does it take a buyer to find a match? How long does it take a supplier to get their first transaction? Set a target and track it weekly. Reducing this number is often the fastest lever for improving liquidity.
• Run a multi-tenanting check on your supply side. Ask a sample of your suppliers directly: what other platforms are you active on? How often do you transact there versus here? This tells you your real competitive exposure, not your perceived one.
• Separate your retention cohorts by join date. Do not look at blended retention. Compare users who joined 6 months ago to users who joined 12 months ago for the same time period post-join. Improving cohort retention over time is the most reliable signal that your network effects are real.
• Distinguish GMV from revenue clearly in your own reporting. Know your take rate, track it over time, and understand whether pressure on it is coming from competition, supplier leverage, or deliberate strategy. Conflating GMV and revenue is one of the most common ways early founders misread their own business health.
• Map your switching costs honestly. Go through the process of signing up and transacting on your top two competitors. Time it. Note what data or preferences a user would have to rebuild. If it takes less than 10 minutes to replicate your core value elsewhere, your moat is thin and retention mechanics (subscriptions, loyalty, data accumulation) need to become a priority.
Source: a16z