Four years ago, in January 2022, the hype around blockchain was at its peak. NFTs of monkey pictures sold for hundreds of thousands of dollars, startups raised huge rounds promising web3 for everything, the metaverse was the inevitable future, and any executive who didn’t mention blockchain in their quarterly was behind the times. Four years later, with the FTX collapse digested, NFTs become a cultural joke, the metaverse forgotten, and the industry reduced to its own real flows, time for an honest inventory. What remained. What worked. And where the blockchain is still a sensible idea in 2026.
What failed with accumulated evidence
NFTs as collectible art are the most visible and most documented failure. Transaction volumes fell more than ninety-five percent from the 2022 peak, dominant platforms have closed or shrunk to maintenance mode, and the highest-cap projects have lost practically all their value. The failure wasn’t technical; it was conceptual. The idea that a public ownership registry of a digital file would give that file value collided with the reality that the file remains infinitely reproducible and the registry only points to a URL that usually goes down over time.
The blockchain-based metaverse is the second clear failure. Decentraland, Sandbox, and the other virtual environments with parcels sold as NFTs had their moment in 2022 and then emptied out. Most report daily active users that would be laughable for any other product; those still operating do so with minimal teams and no growth expectations. The problem wasn’t blockchain itself; it was that the metaverse as a product didn’t work either, and tying the two ideas together dragged both down.
Decentralized applications under the web3-for-everything umbrella, from social networks to email, are the third failure. The promise of replacing centralized infrastructure with decentralized equivalents ran into two hard realities: user experience is worse, and economic incentives to maintain the network tend to be unsustainable once new token buyers stop showing up. What remains of web3 is very small niches with militant users, not real substitutes for centralized services.
What did work
Stablecoins are the big success story. In 2026 they move volumes comparable to traditional payment rails in certain corridors, especially cross-border payments from individuals and SMEs to emerging markets, where banking alternatives are slow, expensive, or outright inaccessible. USDT, USDC, and regulated stablecoins issued by European and American banks under formal frameworks operate with trillion-dollar annual volumes and fees that are a fraction of traditional rails. Blockchain here delivers real value: fast settlement, no middlemen charging to simply move the message, and programmability.
Bitcoin as a digital reserve asset, outside any central bank’s monetary policy, is the second survival story. It’s not used as an everyday payment system outside very specific cases, but it works as a value store in countries with chronic inflation or capital controls, and has consolidated as a recognized category in institutional portfolios. ETFs approved in various markets since 2024 have normalized access. The philosophical debate over whether it’s money or digital gold has lost relevance; it’s an asset class with its own logic.
Traceability in supply chains is the third case where blockchain delivers measurable value, albeit with caveats. Platforms like the retreating IBM Food Trust and its smaller but more specialized successors still operate for sectors where traceability is regulation or competitive advantage. It works when several independent actors need to see the same record without trusting any central one. It doesn’t work when a single actor could use a normal database without complications.
Verifiable digital identities, based on W3C DID standards and verifiable credentials, have matured and are starting to see real deployments in Europe with the eIDAS 2.0 framework and the European Digital Identity Wallet. Although not all implementations use blockchain as substrate, those that do make sense when issuer and verifier don’t share a common trust authority. It’s an area with slow but sustained adoption, without noise.
The case of Ethereum and L2
Ethereum deserves separate mention for how it has evolved. The transition to proof of stake in 2022 eliminated the argument that it consumed electricity like a small city. Layer 2 solutions like Arbitrum, Optimism, and zkSync have turned Ethereum into something viable for low-cost transactions without sacrificing main-chain security. In 2026 Ethereum is the infrastructure where most serious stablecoins settle, where decentralized finance apps with real volume run, and where verifiable identities from several institutional projects are registered.
The DeFi ecosystem around Ethereum shrank significantly during 2022 and 2023, but what remained are platforms with stable volumes, serious audits, and professional teams. Collateralized lending, decentralized exchanges, and derivatives still operate with trillions in annual volume, with occasional crises but without the catastrophic collapses of the previous cycle. It’s not massive but it’s real.
The case of public registries
A case that deserves nuance is verifiable public registries, like property registries, notarial acts, or academic certificates. Several governments have experimented with these uses over the past decade. Results are mixed. They work well when the case requires any party to verify authenticity of a record without going through a third party, and when the government publishing it is one of the parties. They don’t work when blockchain infrastructure adds complexity without delivering measurable benefit over a traditional digital signature with timestamping.
The lesson is that blockchain for public registries only makes sense when radical transparency and independent verifiability are hard requirements. If the case can be solved with a cryptographically signed database, that’s usually the better solution, simpler and more maintainable.
My reading
Blockchain in 2026 is what would have been reasonable to expect a decade ago if the industry hadn’t gone mad in 2021: a technology useful for a handful of specific cases, invisible to the end user in most of them, and unable to replace general infrastructure. Stablecoins and cross-border payments are the star case. Bitcoin as an asset class has its stable niche. Traceability, verifiable identities, and certain public registries deliver measurable value when the problem fits the tool.
Everything else, picture NFTs, the metaverse, web3 as general substitute, enterprise private chains for cases a database would solve better, is history. The 2026 discipline is to always ask: what does this do that a centralized, signed, replicated database wouldn’t do better? If the answer is nothing specific, blockchain probably isn’t the solution. If the answer points to lack of trust between parties, settlement without intermediaries, or independent verifiability, then maybe yes. The industry has learned the hard way to tell the two cases apart, and that’s good news.
