Growth Metrics and Analytics

Web3 Metrics Are Different

Blockchain applications can combine product analytics with on-chain records. An address is not necessarily one person, and activity outside the chain will not appear in transaction data.

Define what each dataset measures: accounts, addresses, transactions, sessions, or people. Avoid combining these as though they were the same unit.

The Core Metrics

1. Total Value Locked (TVL)

TVL is the total dollar value of crypto deposited in a protocol's smart contracts. It is the most-quoted metric in DeFi and the most misunderstood.

What it tells you: The value counted by the chosen deposit methodology. Some assets may already be lent, locked, or counted through related positions, so TVL is not the same as immediately available borrowing liquidity.

What it doesn't tell you: Whether that capital is sticky. During yield farming booms, TVL can 10x overnight when a protocol offers generous token rewards, then collapse to near-zero when those rewards end.

Additional checks: Examine deposits and withdrawals around incentive changes, asset-price effects, and borrowed or recursively deposited assets. Subtracting token rewards from TVL does not isolate organic demand.

Track at: DeFiLlama

2. Unique Active Wallets (UAW)

The on-chain equivalent of MAU. Count the distinct wallet addresses that interacted with your contracts in a given period.

Timeframe	What it measures
Daily Active Wallets (DAW)	Engagement intensity
Weekly Active Wallets (WAW)	Healthy usage cadence
Monthly Active Wallets (MAW)	Broad adoption

Caveat: One person can own many wallets. Airdrop farmers routinely operate 50-100 wallets (sybil activity). Cross-reference with other signals like transaction value and frequency to filter noise.

3. Protocol Revenue

The real fees users pay. Not token incentives, not VC money - actual demand for the protocol's service.

Activity	Revenue question
Swaps	Which fees go to liquidity providers, the protocol, or another party?
Lending	What part of interest is retained, and what costs or losses offset it?
Staking services	What fee is charged on rewards, and who receives it?
Marketplace sales	Which charges are platform revenue rather than seller proceeds?

Track at: Token Terminal

4. Retention

What percentage of wallets that transact in week 1 return in week 2, 4, and 12?

This is harder to measure on-chain than in Web2 (no session tracking), but Dune Analytics queries can approximate it by tracking wallet cohorts over time.

-- Example: 7-day retention for a DEX
WITH first_swap AS (
 SELECT
 sender,
 MIN(DATE_TRUNC('week', block_time)) AS first_week
 FROM dex.trades
 WHERE project = 'uniswap'
 GROUP BY sender
)
SELECT
 first_week,
 COUNT(DISTINCT f.sender) AS cohort_size,
 COUNT(DISTINCT CASE
 WHEN t.block_time BETWEEN f.first_week + INTERVAL '7 days'
 AND f.first_week + INTERVAL '14 days'
 THEN f.sender
 END) AS retained_week_2
FROM first_swap f
LEFT JOIN dex.trades t ON f.sender = t.sender
GROUP BY first_week
ORDER BY first_week DESC

Choose a cohort definition and return action appropriate to the product. Compare periods using the same method; there is no universal retention threshold that establishes product quality.

5. Token Holder Distribution

How concentrated is token ownership? If 10 wallets hold 80% of the supply, governance is effectively centralized regardless of what the documentation says.

Label treasury, exchange, bridge, and vesting addresses before interpreting concentration. One address may represent many users, while one holder may control several addresses.

Tools of the Trade

Tool	What it does	Cost
Dune Analytics	Custom SQL dashboards on blockchain data	Free (paid for private queries)
DeFiLlama	TVL tracking across all chains and protocols	Free
Token Terminal	Revenue, earnings, and valuation metrics	Free tier + paid
Flipside Crypto	Blockchain analytics with bounty programs	Free
Nansen	Wallet labeling and smart money tracking	Paid ($150+/mo)

Key takeaways

• TVL is the most-quoted metric but is easily inflated by token incentives. Always look at TVL alongside revenue.
• Unique Active Wallets is the closest on-chain equivalent to MAU, but is susceptible to sybil inflation.
• Read revenue alongside costs, subsidies, retention, and the definition of the service being measured.
• Retention analysis requires Dune SQL queries tracking wallet cohorts over time.
• Choose data tools based on network coverage, methodology, freshness, and the questions being asked.