Back to all posts

What application revenue tells you about which chains to run infrastructure on

Every infra team ends up asking the same question. Which chains deserve dedicated validators and RPC capacity, and which are noise?

Most teams answer it with the wrong inputs. Install counts, incentivised TVL, and narrative cycles all move first and mean least. Application revenue moves last and means most. This post explains how to read it, and what it signals for infrastructure demand.

What counts as application revenue

Application revenue is what users actually pay to use apps on a chain. It covers trading fees, app take rates, and protocol fees earned by deployed applications. It is different from chain revenue, which is gas paid to the network itself.

Both matter, but application revenue leads. Apps only earn when users transact. Sustained app earnings mean sustained transactions. Sustained transactions mean sustained load on the infrastructure underneath.

Why the usual metrics mislead

Adjust's Finance App Insights 2025 report makes the case in one chart. Crypto app installs grew about 90% year over year. Sessions grew 2%.

That gap is the whole story. Installs measure marketing spend. Sessions measure habit. Revenue measures conviction, because paying users have decided the product is worth real money.

TVL has a related problem. Incentive programmes can rent deposits for a quarter. When emissions stop, the capital leaves and the growth reverses. Fees paid are much harder to rent.

Where the fee activity is right now

Public fee dashboards put Solana's May application revenue at roughly $91M. Hyperliquid apps earned around $53M and Ethereum apps around $52M. Check the current month before quoting these figures, the league table moves fast.

This week's flow data points the same way. Circle minted another $1B of USDC on Solana, taking reported 2026 issuance past $64B. Solana's tokenised RWA market cap climbed past $3B to a record, having added more than $2B in six months. On Ethereum and BNB Chain, Ondo's tokenised stocks went live on Uniswap, listing 430+ assets with a reported $1B+ TVL.

The individual numbers matter less than the pattern they make. Chains where users already pay real fees keep attracting the flows that pay more fees.

What revenue-led chains demand from infrastructure

Fee activity translates into infrastructure load in four specific ways.

Trading apps hammer RPC endpoints. Perp venues and DEX routers poll state constantly and punish slow responses. RPC demand on fee-heavy chains grows with volume, not with headlines. We broke down the cost side in our RPC endpoint hosting comparison.

Revenue attracts institutions, and institutions audit operators. RWA issuers ask about uptime SLAs, key management, and failure domains before they settle real assets. We covered those requirements in our RWA infrastructure piece.

Validator economics hold up after emissions. On revenue-led chains, fee flow supports validator returns without leaning on inflationary rewards. That matters when you are committing hardware for years, not quarters.

Archive and indexing demand compounds. Every trading app needs historical state, and archive nodes are where cloud storage bills climb fastest. Our archive node cost breakdown shows how far that goes.

How we use this signal at LinkPool

We run validators and RPC infrastructure across the major chains on owned hardware. Capacity planning starts with fee data, not announcement cycles. A chain with rising application revenue earns a place on the roadmap. A chain with rented TVL and flat fees does not.

That discipline comes from paying for the hardware ourselves. We have run mainnet infrastructure since 2017, and owned infrastructure makes you honest about where demand is real.

Where to find the data

  • DefiLlama tracks fees and revenue per chain and per app, free.
  • Token Terminal aggregates application revenue and normalises it across chains.
  • Blockworks Research and Artemis publish chain-level fee dashboards with historical trends.
  • Stablecoin issuance is visible onchain, and public trackers report large mints in near real time.

Cross-check at least two sources before committing capacity. Dashboards differ on what they count as application revenue versus chain revenue.

How to read the numbers yourself

  • Pull application revenue by chain from a public dashboard, monthly, same source every time.
  • Compare the fee trend to the TVL trend. Fees flat while TVL grows suggests rented capital.
  • Watch stablecoin issuance as a leading indicator of transactional demand.
  • Check sessions or active-user data where available. Installs alone are marketing.
  • Then put validator and RPC capacity where the paying users already are.

Revenue is not a perfect metric. But it is harder to game than anything else teams use to pick chains. If your infrastructure roadmap ignores it, you are planning around narrative instead of demand.

Frequently asked questions

What is application revenue for a blockchain, and how is it different from chain revenue?

Application revenue is what users pay to use apps on a chain: trading fees, app take rates, and protocol fees earned by deployed applications. Chain revenue is the gas paid to the network itself. Application revenue leads, because apps only earn when users transact, so sustained app earnings signal sustained load on the infrastructure underneath.

Why is application revenue a better signal than TVL or install counts for choosing which chains to run infrastructure on?

Installs measure marketing spend, and TVL can be rented for a quarter with incentives, but fees paid are much harder to fake because paying users have decided the product is worth real money. A chain where users already pay real fees keeps attracting the activity that generates more fees, which is what actually drives sustained validator and RPC load.

How does application revenue translate into infrastructure demand?

Fee activity shows up as load in four ways: trading apps poll RPC endpoints constantly and punish slow responses; revenue attracts institutions who audit operators on uptime and key management; fee flow supports validator economics after token emissions taper; and every trading app needs historical state, which drives archive and indexing demand.

Where can you find application revenue data by chain?

Public dashboards like DefiLlama and Token Terminal track fees and application revenue per chain and per app, and Blockworks Research and Artemis publish chain-level fee dashboards with historical trends. Cross-check at least two sources before committing capacity, because dashboards differ on what they count as application revenue versus chain revenue.