Original author: Sanqing, Foresight News
On April 13, Eli Ben-Sasson, CEO of StarkWare, the ZK-Rollup infrastructure company behind Starknet, announced at an all-hands meeting that the company would undergo layoffs and restructure into two independent business units focused on revenue generation and Starknet development. StarkWare initially launched the StarkEx scaling engine and introduced Starknet to mainnet in late 2021 as an Ethereum Layer 2 validity Rollup, while also developing its own Cairo programming language, Sierra intermediate representation, and post-quantum proof system, establishing itself as a technical benchmark in the ZK-Rollup space. In 2022, the company completed multiple funding rounds totaling approximately $260 million, with its valuation peaking at $8 billion—making it one of the highest-valued ZK projects in the crypto industry at the time.
According to DefiLlama data, Starknet’s on-chain monthly revenue peaked at nearly $6.3 million in November 2023, but since April 2024, its monthly revenue has dropped to only tens of thousands of dollars, a decline of over 95%.
Starknet Chain Fees (Monthly)
Retreat from "infrastructure" to "standalone applications"
In addressing the shift from a "platform infrastructure company" to a "product-driven technology company," Ben-Sasson admitted that StarkWare was previously "too large and too inefficient," and that the company must now return to a startup mode, rapidly iterating with small teams to find product-market fit.
Image source: Eli Ben-Sasson tweet
During this round of industry contraction, StarkWare is not an isolated case. In March, OP Labs (the core development team of Optimism) laid off approximately 20 employees (about 20%) to focus on core priorities, accelerate decision-making, and reduce coordination costs. In January, following an acquisition, Polygon Labs underwent organizational integration, resulting in approximately 60 layoffs across multiple teams; however, the company stated that its net headcount remained unchanged.
Additionally, the exchange Crypto.com laid off 12% of its staff, the L1 Algorand Foundation laid off 25%, and several other companies and projects, including the crypto research firm Messari, have also undergone new rounds of workforce adjustments.
After the reorganization, CFO Ran Grinshtein will oversee back-office functions such as finance and human resources, while the front-end business will be split into two units, each with its own independent BD, engineering, and GTM teams.
- Starknet Development Team: Led by Product Lead Tom Brand, continuing foundational work on the core protocol.
- Product Department: Led by Chief Product Officer Avihu Levy, tasked with direct revenue generation and focused on building products that can only be achieved with the StarkWare tech stack and have minimal external dependencies.
Although the official product lineup has not been disclosed, combined with Levy’s recent paper on implementing quantum-secure transactions (QSB) without modifying the Bitcoin protocol and Starknet’s rollout of privacy features similar to Zcash, quantum-secure products related to Bitcoin are likely to be among their initial areas of exploration.
The Impact of EIP-4844 and the Polarization of L2s
Starknet's predicament essentially reflects the collective discomfort across the entire L2赛道 following protocol upgrades.
In March 2024, Ethereum introduced EIP-4844, drastically reducing the cost of blob data and directly undermining the business model of L2s that relied on profiting from the gas price difference between L1 and L2.
Subsequently, Ethereum has continuously increased blob supply through multiple upgrades; the Pectra upgrade in May 2025 raised the target from 3 per block to 6 per block (with a maximum of 9), and after the Fusaka upgrade at the end of 2025, it was further increased to a target of 14 per block and a maximum of 21 per block.
Ethereum plans to continue gradually increasing blob capacity through additional BPO mechanisms and technologies like PeerDAS, ensuring that L2 data availability costs remain extremely low in the long term.
When the cost of data availability drops significantly, the network's competitive advantage is no longer about being cheap, but rather about its user density and ability to retain capital.
Similarly affected by EIP-4844, the L2 market has shown highly divergent performance.
According to DefiLlama data, Base generated $754 million in revenue in 2025 (62% of all L2 revenue), processing over 60% of total network transaction volume, thanks to Coinbase’s strong user traffic and fiat on-ramp infrastructure. Arbitrum, meanwhile, has maintained its TVL at a stable $2 billion level through a composible DeFi stack anchored by leading protocols such as GMX and Pendle.
Optimism previously relied on the OP Stack and the Superchain ecosystem, but currently, Superchain’s TVL is heavily dependent on Base (over 80%), while OP Mainnet itself accounts for only a single-digit percentage. Its TVL and on-chain fee revenue have also declined significantly between 2025 and 2026. Compounding the issue, Base announced in February 2026 that it would depart from the OP Stack and transition to its own unified tech stack, further undermining Optimism’s pivotal role in the L2 ecosystem.
Left: Pie chart of TVL across Superchain chains | Right: OP Mainnet TVL and Chain Fees (monthly)
Starknet is even more pronounced, with its TVL currently at only about $241 million—less than one-twentieth of Base’s; its native token, STRK, has plummeted from its February 2024 airdrop high to $0.033, with a total market cap of approximately $187 million—lower than the company’s historical funding total of $260 million.
Starknet TVL, STRK Price, and STRK Market Cap
Distribution power determines who stays at the table.
“The infrastructure itself cannot win the race.” This statement by Ben-Sasson reflects on StarkWare’s strategy over the past eight years of “building the network and waiting for users to come.”
StarkWare has invested far more in cryptographic engineering than its peers, developing its proprietary Cairo language and quantum-resistant STARK system from scratch—both highly sophisticated. However, in practice, its refusal to support EVM compatibility has created a formidable barrier for developers, contributing to constraints on ecosystem growth.
The primary drivers of L2 growth are no longer technological differentiation, but distribution capabilities and strategic partnerships. Currently, Base and Arbitrum together lock up nearly 75% of total L2 value.
21Shares predicts that the L2 landscape will consolidate by year-end into “a smaller, more resilient set of networks.” In this winner-takes-all race, retreating to proprietary applications is one of the few remaining differentiators for StarkWare.
Technical expertise is just the entry ticket, not the finish line. What StarkWare now needs to prove to the market is not what cutting-edge technologies it can invent, but what products it can truly sell.