Where NFT activity actually went

The interesting fact about NFTs in 2026 isn’t that interest collapsed. It’s that it collapsed unevenly.

Speculative terms fell hard. Searches asking whether NFTs were dead peaked in 2024 and then declined sharply, which is itself a signal, because people stop asking once they have concluded. Meanwhile searches for NFT utility and gaming NFTs rose.

That isn’t a market recovering. It’s a market changing what it’s for.

How big is the NFT market now?

Total NFT trading volume fell roughly 37% in 2025 to about $5.5 billion. Aggregate market value dropped from around $9 billion to roughly $2.4 billion. Early 2026 saw some value return, but the recovery is concentrated among existing participants rather than new money, and liquidity is poor — the gap between a listed price and an achievable price remains wide across most collections.

If you’re here to evaluate NFTs as an investment, that’s the honest backdrop.

Which NFT use cases actually survived?

Four categories, roughly in order of how well they work.

Do NFT tickets work?

The clearest success, and the one most people never notice.

The problems it addresses are real and predate blockchains: counterfeit tickets, scalping, and no reliable way to verify a resold ticket. A token-based ticket is verifiable, and resale rules can be enforced in code — a price cap, a transfer window, a royalty to the organiser on resale.

It works because the token does a job that’s awkward to do otherwise, and because the user experience hides the machinery. Most attendees have a ticket in an app and no idea a blockchain is involved. That invisibility is a feature; every case where users must understand the technology has adopted more slowly.

Do NFT in-game items work?

More promising in principle than in execution so far, and worth understanding why.

The appeal is straightforward: items you buy in a game are currently entries in a publisher’s database, revocable at any time and worthless if the game closes. Token-based items persist independently and can be sold without the publisher’s involvement.

The catch is that the games built around this idea have mostly been poor. When an economy is the design centre, play becomes a means of extraction, and what emerges resembles a spreadsheet more than a game. Several high-profile play-to-earn titles collapsed when their token economics did, which was predictable — an economy whose rewards depend on new entrants isn’t an economy.

The more plausible direction is established games with real players adding token ownership to items that already matter. That’s slower, less investable, and more likely to work.

Are NFT credentials useful?

Quietly useful. A certificate as a token is verifiable by anyone without contacting the issuer, can’t be forged, and remains valid if the issuing institution changes systems or closes.

The wrinkle is that credentials should generally be non-transferable — a degree you can sell isn’t a degree. This produced “soulbound” tokens, which are permanently bound to an address. That in turn creates a problem: if a credential is locked to an address and the key is lost, so is the credential. Nobody has solved this cleanly.

Can NFTs authenticate physical goods?

Luxury goods, limited editions and collectibles paired with a token recording provenance. Useful in markets with genuine counterfeit problems and where provenance affects value.

The limitation is inherent: the chain can prove the token’s history, but it can’t prove that the token is still attached to the right physical item. A trusted party must make that link, and at that point you’ve a decentralised record with a centralised dependency. It’s an improvement on paper certificates rather than a solution to authentication.

Which NFT use cases failed?

Worth naming directly.

• Profile-picture collections as investments. The overwhelming majority are illiquid and worth a fraction of mint price. Some established collections retain value; almost nothing launched after 2022 did.
• Roadmap-driven projects. Selling a token on promises of future utility. The promises largely went unmet, and in several jurisdictions regulators took an interest in whether these were unregistered securities.
• Play-to-earn as a category. Economics dependent on new entrants funding existing ones don’t survive the entrants stopping.
• The metaverse land rush. Virtual real estate priced on the assumption of imminent mass adoption that didn’t arrive.

How do you tell real utility from a utility label?

Many projects claim utility. The claim is cheap. One question separates them:

Would I want this if I could never sell it?

If the access, the item or the credential is worth having on its own terms, the token is doing real work. If its worth depends entirely on someone paying more later, it’s a collectible with a label, and that’s fine, provided you know which one you’re holding.

Supporting questions:

• Does the utility exist now, or is it promised? Promises have no enforcement mechanism.
• Does it need a blockchain? If a database would serve equally well, the token is decoration.
• Who provides the utility? If a company must stay in business for the token to do anything, you’re trusting that company, which is fine, but it isn’t the trustless story.
• What happens if the project stops? For a ticket, nothing — you attended the event. For a token granting access to a service, everything.

Where does that leave NFTs?

NFTs found narrower, less exciting, more defensible jobs. Verified tickets. Items that persist beyond a publisher. Credentials that check out without a phone call. None of it supports the 2021 valuations and none of it needs to.

For anyone learning this now, that shift is the useful thing to understand. The technology questions, what a token is, how custody works, what a contract can do, are worth learning because they’re stable. The market questions were answered, and not in the direction the courses of that era assumed.

If you’re starting from the beginning, what are NFTs covers the mechanism. If you want to understand the layer below, smart contracts explained is the next step.