Base took 62% of L2 revenue in 2025. Is Coinbase's user funnel breaking the permissionless L2 dream?

General
1

I’ve been working on L2 infrastructure for 6 years now, and the 2025 revenue numbers from Base have me genuinely conflicted about where Ethereum’s scaling roadmap is heading.

The Numbers Are Staggering

Base captured $75.4 million in revenue in 2025—that’s 62% of the entire $120.7 million generated by all Layer 2 networks combined. Let that sink in for a moment. One rollup, owned by a centralized exchange, is generating nearly two-thirds of all L2 revenue.

The growth metrics are even more dramatic:

  • 30x year-over-year revenue growth
  • $4.63 billion in DeFi TVL (46% of the entire L2 market)
  • ~50% of all DEX volume among L2s throughout 2025
  • Aerodrome alone generated $160.5 million (43% of Base’s total app revenue)
  • Morpho’s TVL on Base: 1,906% YoY growth ($48.2M → $966.4M)

When I worked at Polygon and Optimism, we obsessed over performance metrics—transaction throughput, gas costs, finality times. But Base’s dominance isn’t coming from superior technical performance. It’s coming from distribution.

The Coinbase Moat Nobody Can Replicate

Here’s what Base has that Arbitrum, Optimism, zkSync, and every other permissionless L2 can’t match: direct access to 9.3 million monthly active Coinbase users.

The onboarding flow is frictionless:

  1. User already has funds on Coinbase
  2. Click “Bridge to Base”
  3. Funds appear on Base L2 in under 60 seconds
  4. No separate wallet setup, no seed phrases, no bridge risk education

Compare that to bridging to Arbitrum or Optimism:

  1. Download MetaMask
  2. Understand seed phrase security
  3. Buy ETH on exchange
  4. Withdraw to self-custody wallet
  5. Research bridge safety
  6. Wait for bridge finality
  7. Pay gas fees twice (L1 → bridge, bridge → L2)

For the average user, Base removes six friction points that kill crypto adoption. From a UX perspective, it’s brilliant. From a decentralization perspective, it’s terrifying.

Stage 1 Decentralization Isn’t Enough

To Base’s credit, they hit Stage 1 decentralization in April 2025 with permissionless fault proofs and a security council. That’s meaningful progress compared to many L2s still running trusted operators.

But Coinbase still runs the only sequencer. They control transaction ordering. They can theoretically censor transactions. Yes, users can force inclusion via L1, but realistically, how many retail users even understand that escape hatch exists?

The roadmap mentions “third-party sequencers in the future,” but when? And even if they decentralize sequencing, the distribution moat remains. Coinbase will always have the easiest onramp.

Are We Building a Winner-Take-Most L2 Market?

Here’s what keeps me up at night: What happens to permissionless L2s if one exchange-owned chain captures all the users and liquidity?

When Aerodrome generates $160.5 million in revenue on Base, liquidity providers notice. When Morpho’s TVL grows 19x on Base, DeFi protocols notice. The network effects are brutal:

  • Developers build where the users are → Base
  • Liquidity flows where the developers are → Base
  • Users go where the liquidity is → Base

Arbitrum and Optimism are still generating revenue ($23M and $15M respectively in 2025), but that gap is widening. Can they sustain long-term development competing against a chain with a built-in user funnel?

The Uncomfortable Question

I spent years believing Ethereum’s L2-centric roadmap would lead to a diverse, permissionless scaling ecosystem. Multiple rollups competing on performance, UX, and innovation.

Instead, we might be heading toward:

  • One dominant CEX-owned L2 (Base) capturing retail users
  • A few specialized L2s (Arbitrum for DeFi power users, zkSync for privacy-focused apps)
  • Dozens of failed L2s that couldn’t solve distribution

Is this still the Ethereum we were building? Where users have meaningful choice? Where permissionless innovation can compete against entrenched platforms?

Or have we just recreated the traditional tech playbook—where distribution moats beat better technology, and the company with the biggest user base wins?

I don’t have answers. But watching Base capture 62% of L2 revenue while running a centralized sequencer and relying entirely on Coinbase’s brand… it makes me wonder if we’re solving Ethereum’s scalability at the cost of its decentralization ethos.

What do you all think? Am I overthinking this? Or is this exactly the centralization risk we should have seen coming?

2

Lisa, your numbers are spot-on, and honestly, they’re exactly why I’ve been shifting more of our yield optimization strategies to Base over the past year.

Following the Liquidity, Not the Philosophy

Here’s the uncomfortable truth from a DeFi builder’s perspective: I build where the users are. And in 2025, users followed the path of least resistance straight to Base.

When Aerodrome is generating $160.5M in revenue—that’s 43% of all Base app revenue—that tells me where the real liquidity is. When we deployed our yield aggregator on Base vs. Arbitrum, the difference was stark:

  • Base: 8,200 DAU in first month, $12M TVL in 90 days
  • Arbitrum: 1,100 DAU in first month, $2.3M TVL in 90 days

Same product. Same team. Different distribution channel. The Coinbase funnel is real.

The Permissionless Dream Was Always Idealistic

You mention “breaking the permissionless L2 dream,” but I’ll be blunt: most users don’t care about permissionlessness. They care about:

  1. Can I use this without reading a PhD thesis?
  2. Will my funds be safe?
  3. Are the yields competitive?

Base answers “yes” to all three for the average user. The fact that Coinbase runs the sequencer? Most retail users don’t even know what a sequencer is, let alone care who runs it.

We romanticize decentralization in crypto forums, but mainstream users chose Coinbase because it’s centralized and regulated. They want customer support. They want FDIC insurance narratives. They want a company they can sue if something goes wrong.

But Here’s My Concern

That said, you’re right to worry about concentration risk. If Base captures 62% of L2 revenue and continues this trajectory, we face some serious DeFi composability questions:

What happens when one L2 holds all the liquidity?

  • Cross-chain protocols become less efficient (more slippage, worse UX)
  • DeFi innovation gets siloed on Base’s ecosystem
  • Coinbase effectively controls DeFi infrastructure risk

When Morpho’s TVL grew 1,906% on Base, that’s great for Morpho users on Base. But it also means liquidity fragmentation across L2s—which ironically was one of the problems L2s were supposed to solve.

The Market Has Spoken

Look, I’m risk-aware enough to know that putting all our DeFi eggs in Coinbase’s basket is dangerous. But I also run a business. When users overwhelmingly choose convenience over decentralization philosophy, I have to meet them where they are.

Maybe the answer isn’t to fight Base’s dominance, but to figure out how to preserve DeFi composability despite it. Cross-chain messaging, shared liquidity layers, something that lets permissionless L2s interoperate without competing on distribution.

Because right now, competing with Coinbase’s 9.3M user funnel? That’s a losing game for everyone else.

3

Diana, with all due respect, “the market has spoken” is exactly the kind of capitulation that undermines everything we’ve been building since Bitcoin’s genesis block.

This Is Exactly What Ethereum Was Supposed to Prevent

Lisa’s concern isn’t overthinking—it’s the central tension of our entire ecosystem. We spent a decade building censorship-resistant, permissionless infrastructure specifically to avoid recreating the centralized chokepoints of traditional finance.

And now we’re celebrating that one centralized exchange controls:

  • 62% of L2 revenue
  • The only sequencer on the dominant L2
  • Transaction ordering for nearly half the L2 market’s TVL
  • The primary onramp for retail users to “decentralized” applications

How is this different from PayPal processing payments for merchants? Sure, it’s built on Ethereum settlement, but if Coinbase controls the user experience and can theoretically censor transactions at the sequencer level, what have we actually decentralized?

Stage 1 Isn’t Nearly Enough

Yes, Base achieved Stage 1 decentralization with permissionless fault proofs. That’s a security improvement—users can force transaction inclusion via L1 if censored.

But let’s be realistic about retail behavior:

  • How many Base users even know their funds are on an L2?
  • How many understand forced inclusion via L1?
  • How many would pay L1 gas fees (potentially -200) to bypass sequencer censorship?

For 99% of Base users, Coinbase is a trusted intermediary with effective veto power. That’s not decentralization theater—it’s just theater.

The promise of third-party sequencers “in the future” rings hollow when Base is already capturing the entire market. By the time they decentralize (if they do), Coinbase will have a network effect moat that makes competition impossible.

The Network Effect Death Spiral

Diana, you say “I build where the users are,” and I understand the business logic. But that mentality creates a self-reinforcing centralization spiral:

Phase 1: Coinbase uses CEX distribution to bootstrap Base
Phase 2: Developers follow users → deploy on Base
Phase 3: Liquidity follows developers → DeFi concentrates on Base
Phase 4: More users follow liquidity → back to Coinbase funnel

The result? Permissionless L2s can’t compete, even if they have:

  • Better decentralization (Arbitrum, Optimism)
  • Superior technology (zkSync, Starknet)
  • More innovative features (Scroll, Polygon zkEVM)

None of that matters when Coinbase controls the onramp. We’ve just recreated Web 2.0’s aggregation dynamics—the platform with the largest user base wins, regardless of technical merit.

This Isn’t Inevitable

Here’s what frustrates me: This outcome wasn’t inevitable. Ethereum had alternatives:

  • Enshrined rollups with neutral sequencing
  • Shared sequencer networks (like Espresso or Astria)
  • L1 improvements that made L2s less necessary
  • Better decentralized bridge UX to compete with CEX funnels

Instead, we got a fragmented L2 landscape where the CEX-owned chain wins by default because it has the easiest onboarding. That’s a coordination failure, not a market verdict.

What’s At Stake

If Base’s dominance continues, we’re headed toward:

  • Regulatory capture: Coinbase must comply with US regulations. That means KYC, transaction monitoring, potential censorship of sanctioned addresses at the sequencer level.
  • Single point of failure: When (not if) Coinbase faces regulatory pressure or technical issues, nearly half of Ethereum L2 activity goes down.
  • Vendor lock-in: DeFi protocols that build on Base become dependent on Coinbase’s infrastructure decisions, governance, and fee structures.

Diana, you mention “figuring out how to preserve DeFi composability despite Base’s dominance.” But composability assumes neutrality. If Coinbase controls the dominant platform, they control the standards, the messaging layers, and the interoperability choices.

That’s not DeFi composability—it’s building on someone else’s platform and calling it decentralized.

We Can’t Outsource Decentralization

Look, I’m not naive. I know retail users want convenience. I know Coinbase’s onboarding UX is better. I know distribution matters.

But if the crypto ecosystem’s answer to “how do we scale Ethereum?” is “let centralized exchanges run the L2s,” then we’ve fundamentally lost the plot.

Lisa asked if she’s overthinking this. She’s not. This is the exact centralization risk we should have seen coming—and we still have time to address it before it’s irreversible.

The question isn’t whether Base will dominate short-term. It’s whether we’re willing to accept that dominance as permanent, or if we’ll build alternatives that actually preserve the decentralization properties that made Ethereum worth building in the first place.