When a tool resale marketplace opened in Manchester in 2022, it launched with 1,400 registered sellers and a write-up in the local paper. Eighteen months later it carried 200 active listings a month and spent £11,000 a month to keep going. The founder, who asked to be identified only as Dan, sold the code to a competitor for £4,000. “Year one is the launch,” he said. “Year two is when you find out if anyone comes back.”

The pattern has a shape. A marketplace needs enough listings that a buyer finds the thing they came for, and enough buyers that a seller gets an answer within a day. Economists have studied this configuration for decades under the name two-sided market, and the practical problem beneath the theory is stubborn: supply and demand have to arrive close together, or both sides drift away.

Liquidity is a small number

Operators talk about liquidity as if it were a river. On the ground it is a count. Dan’s marketplace needed roughly 300 live listings in a category before buyers returned weekly. In its best category, cordless drills, it peaked at 190. “Buyers would search, see one page of results, and leave,” he said. “Then sellers would see zero views and stop listing. Both sides were right.”

That is the chicken-and-egg problem in plain clothes. Sellers on the supply side wait for demand, buyers wait for supply, and a small marketplace lives or dies on which side it can subsidize first. Dan paid sellers £5 per listing for three months. It filled categories fast and emptied them in January, when the payments stopped.

Dan saw the same drift on the seller side. A listing that drew zero views in its first week was almost never renewed, and renewal fees paid his hosting. With 260 repeat sellers, the marketplace had 260 customers. Everyone else was passing through.

Where the second year goes wrong

Year one runs on novelty and founder money. Friends list items, local press links to the site, and the graph climbs. Year two runs on habit, and habit is where weak marketplaces lose. A site operator in Ohio who asked not to be named described the drop in her handmade-goods marketplace: “Month fourteen, listings fell 40 percent. Nothing broke. People just went back to Facebook Marketplace, where their friends already were.”

Retention numbers decide it. Of the 1,400 sellers who registered on Dan’s site, 260 listed more than once. Of the buyers who made one purchase, 8 percent made a second one. A marketplace cannot pay its costs on a single transaction per customer, so the second year is really a test of whether the first year built a habit.

What the survivors measured

The marketplaces that made it past year two picked a narrow lane and tracked one number. A motorcycle-parts site in Rotterdam counts parts matched per week. A farm-equipment listing site in Saskatchewan counts phone calls between buyer and seller. Both numbers sat at the center of every decision their owners made.

The Rotterdam owner checks category counts every Monday morning and shifts featured slots toward whatever runs thin. Empty shelves get attention first, before marketing and before features.

Neither business raised money. Both charge small fees (3 to 5 percent on completed sales) and both treat empty categories as an emergency. “If a buyer finds nothing in the first three clicks, I have a week to fix it,” the operator in Ohio said.

Dan’s old marketplace redirects to a holding page now. The domain still gets 400 visits a month from buyers searching for drills. He checks the analytics out of habit. “The demand was real,” he said. “I just couldn’t get enough of it in one place at one time.”