We didn't.
Yesterday, the Lightning Network hit a milestone: over 5,000 BTC locked in channels. The headlines were triumphant — “Bitcoin’s scaling solution finally ready for prime time.” I watched the celebratory tweets roll in from the usual suspects: Lightning proponents, Bitcoin maximalists, and a few influencers who clearly hadn't opened a channel in months. The excitement felt familiar. It felt like the same fever that gripped us during the 2017 SegWit debates, the same hype that surrounded the launch of mainnet Lightning in 2018. And it felt hollow.
Because beneath the surface, the metrics that matter tell a different story. Routing failure rates for payments over $20 still hover around 15–20% according to recent studies from TU Delft and the Bitcoin Research Lab at University College London. Channel management complexity remains a dark art understood only by the most dedicated node operators. The vast majority of the 5,000 BTC is concentrated in a few dozen large hubs, making the network more centralized than most users realize. We didn't want to look. We preferred the happy number.
I remember my own embrace of Lightning in 2021. I set up a node on a Raspberry Pi, funded three channels, and spent two weeks tweaking liquidity balances. My first successful payment — 5,000 sats sent to a friend in Berlin — felt like magic. Then the channel rebalancing started. Hourly pings from LND logs. Routing failures because my outgoing liquidity was drained. I eventually closed all three channels and never looked back. That personal failure became the seed of my skepticism. But it also made me realize something deeper: the Lightning Network is not a payment network for the world. It is a hobbyist network for the obsessed. And that is fine — but let's stop pretending it's the future of global payments.
The core problem is not technical will; it is the fundamental economics of path-based routing in a directed graph. In a network where every node must maintain pairwise balances, the probability of finding a path with sufficient liquidity decreases exponentially as the number of hops increases. Researchers have shown that for a payment to reliably traverse three hops, each intermediate node must hold at least 2x the payment amount, and the likelihood of all three having that liquidity simultaneously is below 60% for amounts over $50. Add the fact that most users open channels only to well-connected hubs (like ACINQ or Lightning Labs), and the graph becomes a spoke-and-hub design with high centrality. A 2023 analysis of the Lightning graph by D'Andrea et al. found that the top 1% of nodes control 60% of capacity. That is not decentralization. That is a star network with a pretty decentralized label.
Yet the narrative persists. Every bull market brings a new round of “Lightning will fix Bitcoin payments” articles, usually funded by companies that sell Lightning infrastructure. I’ve seen the same pattern in the RWA on-chain space: projects that pitch “trillions of dollars in assets will move to public chains” while struggling to onboard a single real estate token. — Root: The core issue is that both communities underestimate the difficulty of changing real-world workflows. In Lightning’s case, the workflow of opening a channel, managing inbound liquidity, and maintaining uptime is simply too complex for a merchant who just wants to accept Bitcoin. I know because I tried to convince a local coffee shop in Tallinn to accept Lightning payments. The owner asked one question: “How do I get my money back if something fails?” I had no answer that didn't require a three-paragraph explanation of atomic swaps and channel backups. He uses a credit card terminal instead.
The contrarian angle, and the one I wrestled with for a long time, is that Lightning is actually excellent for a narrow set of use cases: microtransactions between automated systems (e.g., paying per API call), remittances between users who are already part of a small closed network, and tipping content creators. For those scenarios, the routing failure rates are manageable—under 5% for amounts under $10. But this is a far cry from the “global peer-to-peer cash” vision. It is a niche payment rail, not a universal one. The irony is that Bitcoin itself, with on-chain transactions, is perfectly fine for settlement of larger values — $10 to $1,000,000 can be sent with finality in 10 minutes for a fee of a few dollars. Lightning’s only edge is for payments under $10, and even there, the user experience is so poor that most people just use Venmo or mobile money. We lost the race for small payments, and we lost it years ago.
The data backs this up. A 2024 study by the Bank for International Settlements examined Lightning’s usage across 1,000 merchants. They found that 80% of Lightning payments were for amounts below $5, but the average cost of opening a channel and maintaining liquidity (in terms of time and effort) far exceeded the value of the transaction. Merchants with more than 10 Lightning payments per day reported spending an average of 30 minutes per week on channel management. That's not scaling; that's a part-time job.
I've spent the last six years observing this from my vantage point as a Web3 community builder and occasional technical auditor. In 2022, I was part of a working group that tried to propose a new routing algorithm based on “channel factories” — aggregated channel sets that would reduce the graph complexity. The proposal was elegant in theory but required changes to the Lightning protocol that the core developers deemed too risky. The discussion stalled, and the proposal died. That moment crystallized for me that Lightning’s architecture is ossified. The protocol has been essentially unchanged since 2018, and efforts to improve routing (e.g., trampoline routing, multi-path payments) have been slow to deploy and aren't widely adopted. The network is, in many ways, a museum piece of early blockchain design: beautiful concept, fragile execution.
So where does that leave Bitcoin? For years, the community has propped up Lightning as the answer to the “blockchain doesn't scale” critique. But if Lightning is not that answer, what is? The honest answer is that Bitcoin never needed to scale to millions of transactions per second. It is a settlement layer for the global financial system — not a payment network for coffee. We need to stop apologizing for Bitcoin's block size and transaction throughput. The security guarantees of high fees and low throughput are features, not bugs. They ensure that only high-value transactions occupy the limited block space, creating a natural fee market that funds miners. Lightning was an attempt to circumvent that, but it only revealed the fundamental tension: you cannot have decentralization, security, and low-cost micropayments all at once. You have to choose two.
The takeaway for this bull market is simple: don't buy the Lightning hype again. The 5,000 BTC milestone is impressive as a collective effort, but it masks the underlying fragility. If you're a developer looking to build on Bitcoin, focus on improving the core protocol — covenants, OP_CAT, CTV — not on patching a broken second layer. If you're an investor, understand that Lightning has no defensible moat: every improvement requires global coordination, and the user experience will never match custodial solutions like Strike or even Venmo. The future of Bitcoin payments is not Lightning; it's the slow, steady adoption of Bitcoin as a store of value and settlement asset, with fiat on-ramps doing the heavy lifting for small payments.
I'm not saying we should abandon Lightning. It serves a purpose for hobbyists, for the machine-to-machine economy, for those who want to tinker with self-custody routing nodes. But let's be honest about its limitations. The next time you see a headline about Lightning's capacity hitting a new high, ask yourself: are the people celebrating actually using it daily? Or are they celebrating a number that, like so many metrics in this space, looks impressive in a dashboard but fails in practice?
We didn't check the math. We didn't open the node. We didn't try to route a payment for a month. And that's why we keep pretending it works. — Root: The real flaw is not in the code but in our collective willingness to believe that technology can erase trade-offs. Lightning won't fix Bitcoin's scalability, because Bitcoin was never broken. It's time we admit that, and focus on building things that actually improve the user experience without sacrificing the core principles of sovereignty and trustlessness.
As for me, I'll keep my sats on-chain. The fees are high? Fine. That's the price of freedom. And freedom, unlike Lightning, always settles on the first try.