If you’re building a data or information business — or running any “sell the shovels” play — this lesson lands on a conclusion that should sting a little: in crypto data, influence is worth more than subscription revenue, because subscription revenue was never worth that much to begin with.

Start with the two defining traits of the “shovel seller” business.

First: resilient in a bear market, capped in a bull market. Data subscriptions and API revenue hold up better than trading protocols do when the market turns down — demand for analysis actually rises as prices fall. But that same business can never capture the direct upside of a bull market the way an exchange does. Bear-market resilience, bull-market underperformance — that’s the fate of every shovel seller.

Second, and far more counterintuitive: openness as a moat. DefiLlama built its de facto standard status with a move nobody saw coming — going completely, permanently free. In an industry where everyone else builds walls by charging money, it became the data source the whole industry defaults to citing by charging nothing. Its moat was never “you can’t afford this.” It’s “you can’t function without this.”

Now look at what happened to the paid camp. 2026 delivered the verdict in two brutal numbers.

• Messari, the standard-bearer of paid research subscriptions, was valued at roughly $300 million four years ago. In June 2026, Blockworks acquired it — reportedly for only around $10 million. A roughly 97% collapse in four years.
• Nansen, the face of on-chain address-labeling analytics, cut its flagship subscription from about $1,299 a month to about $49 a month — the same functionality, a 95% price cut.

Both point to the same underlying disease: on-chain data is inherently a semi-public good. The raw data is available to anyone — that’s a defining feature of blockchains. What you can actually charge for is processing, cleaning, and interface convenience — and that layer has almost no moat. DefiLlama aggregates it for free. AI tools keep driving the cost of processing toward zero. When the raw material is free, a processing-fee business has its ceiling built in from day one.

Which reframes DefiLlama’s strategy entirely. It’s not that it “hasn’t figured out how to charge yet.” It may have simply been right all along: in this industry, influence is worth more than subscription revenue — because subscription revenue was never worth that much.

If you run any data or information product, sit with this one: what are you actually selling? If it’s “a processed version of raw material anyone else can get for free,” your pricing power was never yours to begin with.

Free and open to build influence, or paid and closed to build revenue — if it were your call, which one would you bet on?

— Adapted from Crypto Sector Leaders, Chapter 17: Crypto Data Analytics and Research Platforms

#DataBusinessModels #DeFiLlama #CryptoInfrastructure