[Digital Today reporter Jinju Hong (홍진주)] Two cryptocurrency-based prediction market startups announced they are ending their businesses just 90 minutes apart. The market itself is growing fast, but an analysis says survival is getting harder for late entrants as trading concentrates on Kalshi and Polymarket.

On Aug. 10 local time, blockchain outlet The Defiant reported that prediction market startups Trepa and Fireplace said they would end their services and told users to withdraw funds by Sept. 30.

The two companies had different business models, but the outcome was the same. Total trading volume over the past 30 days across 40 prediction markets tracked by DefiLlama reached $14.87 billion, but Kalshi and Polymarket accounted for $13.87 billion of that. Their combined share reached 93.3%.

Kalshi's share alone was an overwhelming 75.7%. The share of Opinion, a BNB Chain-based prediction market in third place, was 2.8%. That shows a structure in which the market is growing while trading liquidity concentrates even more on a small number of platforms.

Trepa promoted its own Solana-based prediction market structure. It divided rewards based on how close a user's forecast came to the outcome, running 60-second rounds over a 1-hour period each day targeting the bitcoin price. But its founders judged that the structure itself was an obstacle to scaling the business. They said retention worsened the longer users used the service, and they could not find a way to fix the structural problems while expanding the service.

It also struggled to attract users. Its referral program did not work as expected, and paid influencer marketing also had limited results. Partnerships to provide a prediction market widget to outside services were also constrained because the market was open for only 1 hour a day.

Trepa also questioned the size of the potential user base for crypto prediction markets. Its founders thought daily and monthly active user figures for crypto that they had previously referenced might have been inflated. Under their working hypothesis, actual active users for related products were in the hundreds of thousands, not millions. They also cited regional service restrictions and deposit procedures as factors that made it harder to secure early users.

Fireplace took a completely different approach. Instead of building its own order book, it operated as a professional trading terminal on Polymarket run by Hong Kong registrant Enclave HK Limited. It routed users' orders and collected fees.

But this model also faced intensifying competition. Kalshi launched a free desktop terminal, 'Kalshi Pro', on July 13 in the professional trading terminal segment Fireplace targeted.

Kalshi Pro offers a screen that curates about 2,000 markets and a perpetual futures trading interface. With more than 20 interface wrappers also emerging that connect orders to Polymarket, competition in professional trading tools has also intensified.

Fireplace did not disclose specific reasons for the shutdown. It is directing users straight to Polymarket through its service-termination page. It said existing positions also continue in the same Polymarket markets.

A string of closures among prediction market startups does not mean a downturn for the overall market. Instead, funding is flowing to some firms. ProphetX raised $35 million on July 28, and Pascal secured a $9 million Series A investment on July 16. Both firms are focusing primarily on the regulated U.S. market rather than a proprietary on-chain structure.

Trepa's founders also did not deny the potential of the prediction market idea itself. They judged there is still business value in the concept that people might pay to measure how close their predictions are to actual outcomes. But they acknowledged they could not find a way to make it a sustainable business.

The cases show the center of gravity in prediction market competition shifting from unique product structures to securing trading volume, liquidity and user distribution channels. That is because platforms that attract market participants accumulate more trades and liquidity, creating a virtuous cycle that draws in users again.

In the end, even if the prediction market expands, it is becoming a structure in which it is difficult for all operators to grow together. With trading volume continuing to concentrate overwhelmingly on Kalshi and Polymarket, a key variable in future market reshuffling is what differentiation strategies late entrants can use to secure liquidity and users.