Why is time to first earning the metric that matters for supply-side activation?

Because the supply side is running an experiment with their own effort as the stake, and the payback is uncertain. Every day between signing up and earning is a day the platform is asking for unpaid work on a promise. Time to first earning is the only supply metric that predicts whether a creator or seller is still there in 90 days, and it is the one number that improves listings, retention, and word of mouth at the same time.

Why do referral and viral loops fail in marketplaces?

Because a loop is a multiplier on whatever the product already does. Build one before activation works and you accelerate the rate at which people arrive, fail to reach value, and tell others it did not work for them. The honest sequencing is to fix the value moment first, measure whether activated users refer at a materially higher rate than the rest, and only then instrument the loop, since a viral coefficient calculated across unactivated users is a number that describes nothing.

Time to First Dollar: Two-Sided Marketplace Activation

Q: What is activation in a two-sided marketplace?

Two separate events on two different clocks. On the supply side, activation is the first earning: the moment a seller, creator, or provider receives money from a stranger rather than from a friend they recruited. On the demand side, it is the first satisfying transaction, meaning the buyer got what they came for and would repeat it. A single activation rate averaged across both sides describes neither, because the 2 populations churn for opposite reasons and respond to opposite fixes.

What is activation in a two-sided marketplace? Two separate events, on 2 different clocks. On the supply side, activation is the first earning: the moment a seller, creator, or provider receives money from a stranger rather than from a friend they recruited themselves. On the demand side, it is the first satisfying transaction, meaning the buyer got what they came for and would do it again.

One activation rate averaged across both sides describes neither. The 2 populations churn for opposite reasons and respond to opposite fixes, and the supply number is the one that compounds, because a seller who earns produces listings, referrals, and inventory the demand side can actually convert on.

Single-sided products have 1 activation question: did the user get the value they came for. Marketplaces have 2, and they run on different clocks, in different currencies, with different failure modes. Most marketplace dashboards flatten that into 1 number and then wonder why the number will not move.

Why does a marketplace have 2 activation problems?

The demand side is spending money to save time. Their activation is fast, often inside a single session, and their failure mode is not finding what they wanted. The supply side is spending time to make money. Their activation is slow, measured in days or weeks, and their failure mode is effort that never paid.

Those are not 2 versions of the same funnel. A buyer who has a bad first experience leaves quietly and might come back when they need something. A seller who puts in 3 weeks and earns nothing leaves loudly, tells people, and does not return. Same platform, opposite economics of disappointment.

This is the multi-persona problem from B2B with the stakes rearranged, and the structural fix is the same one I described in The Buyer, the Admin, and the User: detect who you are talking to at entry, branch the first session, and stop reporting 1 average that belongs to nobody.

Time to first earning is the number that compounds

Of every metric a marketplace can pick, time to first earning is the one worth building the organization around. It is the supply side's entire question, asked in 4 words, and it predicts 90-day retention better than volume, engagement, or profile completeness ever will.

The reason is uncomfortable. Between signing up and earning, the platform is asking someone to work for free on a promise. Every day in that gap is a day the promise looks worse. Shorten the gap and you are not improving a funnel step, you are reducing the size of the bet you are asking a stranger to place.

It also compounds in a way demand-side metrics do not. A supplier who earns stays, and a supplier who stays produces more inventory, more listings, and more of the thing the demand side is converting on. A buyer who converts produces one transaction. The asymmetry is why supply-side activation deserves the roadmap slot, and it is the same argument for picking metrics that change behavior instead of metrics that are easy to count, which I made in Metrics That Move Teams.

What actually shortens it

Almost never the onboarding flow, which is where teams look first. The gap is usually made of 4 things: the supplier does not know what a good listing looks like on this specific platform, the payout mechanics are unclear enough to feel risky, discovery buries new supply under established supply, and nothing tells the supplier they are close.

The last one is the cheapest and most neglected. A new supplier sitting at zero has no idea whether they are 2 days or 2 months from earning, and in the absence of a signal they assume the worst and stop. A single honest progress indicator, built from what actually correlates with first earning on your data, does more for supply retention than another onboarding step ever will.

The discovery point is the one most platforms get wrong on purpose. Ranking purely on performance is correct for revenue this quarter and fatal for supply growth, because new supply can never accumulate the performance that would rank it. Some deliberate exposure budget for unproven supply is not charity. It is the cost of having supply next year.

Why do referral and viral loops fail here?

A loop is a multiplier on whatever the product already does. Build one before activation works and you accelerate the rate at which people arrive, fail, and tell others it did not work for them.

Viral coefficients calculated across a whole signup base are close to meaningless for this reason. The number worth computing is narrower: do activated users refer at a materially higher rate than everyone else, and do the people they bring activate faster than cold arrivals. If both are true, you have a loop worth instrumenting. If neither is, you have a referral program that will move a vanity number and nothing under it.

Sequencing beats cleverness here. Fix the value moment, prove the referral differential exists among activated users, then build the mechanism. Doing it in the other order is how marketplaces spend a quarter shipping invite flows that amplify churn.

What to instrument

Two funnels, reported separately, always. Supply: signup to first listing, first listing to first transaction, and the distribution of time to first earning rather than its average, because the average hides the tail where all the churn lives. Demand: session to first satisfying transaction, and repeat rate inside 30 days.

Then 1 joint number that tells you whether the sides are in balance: the share of demand searches that end without a match. That is the number that says whether your problem is supply, discovery, or neither, and it is the earliest warning that a growth push on one side is about to break the other.

None of this is exotic. It is ordinary activation work applied honestly to a product where 2 different people have to succeed before either transaction exists, and where the cheap version of measuring it produces a number that sounds fine while the supply side quietly walks. The instrumentation has to exist before the churn does, which is the argument in Building Retention Before You Need It, and it is worth more in a marketplace than anywhere else, because a supplier who leaves takes their buyers with them.