Most investors learned the wrong lesson from the failed consumer startups funded between the NFT Boom in 2021 and the end of memecoins in 2025. They look at that cycle’s consumer crypto experiments and conclude that “non-financial crypto” (the term an oxymoron in itself) fails as a category. But the takeaway that “nobody” wants consumer crypto apps amongst a majority of crypto VCs conflates failed timing for failed demand, and ignoring consumer crypto apps today would repeat a mistake many VCs came to regret in the mid-2000s.

Thomas Aquinas helped give the seven deadly sins their durable moral architecture, and viewed them less as random vices than as recurring distortions of human desire. That same framework maps surprisingly well onto the consumer apps of Web2: pride becomes the profile and follower count, envy becomes the feed, lust becomes the swipe, etc. The old taxonomy persists because the underlying machinery has not changed; Web2 just turned those impulses into different views of status.

And crypto does status exceptionally well: proof of ownership and being early are hallmarks of the technology and the communities around it. It stands to reason that status would be amplified by new consumer crypto apps built around it, but that is not yet true in 2026. Though past consumer crypto experiments were noble attempts to build new durable consumer habits, the infrastructure, UX standards, and cultural conditions were not ready to make them signals that led to big businesses.

But today the rails are finally smooth and cheap, regulation is clear, builders have new AI-powered tools, and agents bring founders an entirely new user group to build products for. Polymarket and Kalshi are two of the pioneers, and crypto’s consumer social cycle is upon us. It follows a historical pattern we saw before the consumer explosion of 2006.

A History Lesson in Pipes

By the late 90s, Pew estimates that 41% of US adults were online. The internet was still slow, clunky, and almost impossible for hobbyist builders to build on, but it had shown enough flashes of magic to be interesting: email, search, online commerce. Utilities in some sense.

Investors noticed and poured hundreds of billions of dollars into telecom companies, laying the fiber needed to improve online experiences. The thinking was that those who owned the rails would own the value created on top of them.

They were wrong. Overbuilding drove down bandwidth prices and made the infrastructure layer a commodity. The pipes were essential to progress—there was a boom and crash of startups experimenting with the internet around the early 2000s—but the pipes themselves were a bad investment. Global Crossing, for example, spent $15 billion laying fiber in 1997, never turned a profit, and went bankrupt in 2002.

Carlota Perez created a framework to describe this moment, and understanding it matters for crypto. The early 2000s marked the completion of the web’s “installation phase.” This is when capital floods into the enabling infrastructure of a new technology, often overshoots the buildout, and leaves behind cheap capacity for the next wave of builders. This following phase is often known as a golden age.

By the mid-2000s, the next wave of builders did come—YouTube, Facebook, Twitter—and the consumer internet as we know it today was made possible by cheap broadband laid out half a decade before.

The investors who bet on the application layer made out with some of the best vintages in venture capital. An important lesson from them is that the first wave did not reveal a ceiling on demand, but instead immaturity of the medium.

Crypto’s Installation Phase

The analogy to crypto is clear, but the point is simple: it is time to pay attention to consumer opportunities because crypto just went through its own “installation phase.” Crypto’s 1998 moment began around 2020, when hundreds of thousands of users were using Ethereum to do things like DeFi that felt like clunky magic. More experimental use cases began to follow.

While everything in crypto is inherently financial, “non-financial” use cases, defined by no immediate cash flows, started to show real signal by 2021. According to Blockworks, NFT communities went from ~85m of volume in 2020 to ~20b in 2021. DAOs and the earliest permissionless social networks became places where builders could connect and experiment. The common thread was status that came with the proof of belonging to something early.

As in the early 2000s, investors poured money into the infrastructure to make these experiments better, and, as in the crash of ’99, most of the experiments during that buildout failed. Farcaster, social tokens, and DAOs were all tried during the buildout of this infrastructure in the early 2020s, not after it had matured into what it is today. Their failure should be read less as a verdict on demand than as evidence that the experiments arrived before the conditions for success did.

AI Provides a New Catalyst for Consumer Experimentation

Finally, it’s not just in crypto that consumer apps have felt stale and unoriginal in recent years. There is a real argument that most possible consumer behaviors for feeds and mobile were already covered during the Web 2.0 mobile era. Adding blockchain to messaging, social networks, and games often resulted in a worse imitation of what users were already using.

But AI changes the premise. For the first time in years, ordinary users can do fundamentally new things with software: talk to it, delegate to it, create with it, personalize it, and build through it. There are ideas like DAOs, tokenized communities, and microtransactions that need to be tried again on our now polished rails. We may even find that agents are the target customers.

At the same time, the builder stack is getting good enough for small teams to ship ideas. Open-source models and cheaper inference make intelligence easier to access. Companies like Uniswap and Coinbase are building agent-native SDKs to make it easier to build onchain apps with agents. We are set up for a Cambrian explosion of experimentation.

The most interesting consumer apps may start weird, niche, or toy-like. Most vibe-coded apps will not become massive companies, but that may matter less if crypto becomes a primitive underneath them. Crypto is unusually good at monetizing participation inside small, passionate networks.

In closing, I built non-financialized consumer products for years, starting in 2022, with Primitives. The app made creating UGC NFTs from your camera roll simple and free. Our core audience minted 500k+ NFTs. It was a passionate community of artists who had never used NFTs before. They were attracted by the idea of strangers “collecting” their artwork and found it more validating than a “like” on Instagram. That was status, not speculation.

The lesson is not that Primitives would have succeeded today; it would not have. It is that we had to overcome many hurdles then that no longer exist just to see whether users wanted our product to begin with. I had to raise millions of dollars, build embedded wallet infrastructure from scratch, and create bespoke logic around Solana congestion issues that simply don’t happen anymore. We were not allowed in the App Store at the time, and Apple Pay did not allow crypto apps. It took us 1.5 years to run our experiment end-to-end.

Today, there are APIs for all of these features, and I could build a fleshed-out version of the app by myself in a week. We are in a crypto winter, total market cap has fallen ~48% from its $4.27T peak in October 2025, but just as in the mid-2000s, the builders now have everything they need to build consumer status businesses.