After a new chain's mainnet goes live, the core problem of "nobody using it" or "rapidly cooling off" is often not a lack of capital (TVL), but that users come and go (retention). TVL can be quickly inflated through subsidies and incentives, but retention rate directly reflects genuine demand and long-term health—it is the signal that exposes problems earlier than TVL.
1. First, Understand the Two Concepts: The Difference Between TVL and Retention Rate
TVL measures "how much money is locked"—it can be rapidly pushed up by offering incentives and airdrops, but that money can be withdrawn at any time. Retention rate measures "whether users will come back"—it is much harder to fake because it reflects real use value and stickiness.
The typical lifecycle of a new chain is: before mainnet launch, attention gathers around funding and narratives; after launch, airdrop incentives trigger a surge in TVL and address numbers. But once incentives taper off, speculative capital and "airdrop hunters" quickly flee, leading to a cliff-like drop in transaction volume and activity.
2. A Classic Case: Berachain's "Cliff" Decline
Berachain was one of the most representative new chains in the first half of 2025. Its innovative "Proof of Liquidity" consensus attracted over $3.1 billion in pre-launch liquidity before mainnet. Barely 20 days after mainnet, TVL briefly hit $3.27 billion, ranking sixth among all chains by TVL.
But from March to June, TVL plummeted 68% to $988 million, and transaction volume in May dropped 87%. Users and capital were highly dependent on airdrop- and reward-driven liquidity, not on sustained organic demand.
PANews' 2025 year-end data confirms this: Berachain had just 18,800 daily active users at year-end, Movement had 12,600, and Scroll only 2,900. These chains did not lack money—they lacked the ability to retain people.
3. Why Retention Exposes Problems Earlier Than TVL
Retention reveals "whether users stay after they come," while TVL only tells you "how much money has entered."
In the first half of 2025, the common pattern among new chains was: many chains quickly entered a consolidation phase after the initial hype—a classic cooling-off after "token launch frenzy." Unichain, AbstractChain, Ink and others all showed bottlenecks in user retention and activity, with transaction volume and fee data visibly weakening, highlighting the mismatch between incentive models and actual application demand.
Ethereum L2 TVL fell to a two-year low in July 2026, but that does not mean L2s have no users—many chains still see high transaction volume. The problem is that "capital no longer has confidence to remain there." User activity can stay lively while TVL declines; this divergence is exactly a direct reflection of the "retention problem."
The average retention rate of the global Web3 network is only 5.4%, while the Web2 benchmark is between 25–40%. When a new chain's retention rate is far below the industry average, a collapse is only a matter of time, even if TVL is still high.
4. Hands-On: How to Use On-Chain Data to Assess a New Chain's Health
Step 1: Check the chain's retention rate data
What to do: Search for "user retention" or "Retention" dashboards on data analytics platforms like Dune or Artemis for the chain.
How to do it:
Find the "30-day retention rate" or "90-day retention rate" data. If the retention rate is well below 20%, it means users come and go. Reference values: Ethereum Q1 2026 retention rate was 26.2%, BNB Chain 20.5%, Base 17.3%, Arbitrum 16.6%, while Solana was only 7.9%.
Distinguish between "activity retention" (whether users continue to transact) and "holding retention" (whether users continue to hold assets); the two reflect different dimensions of health.
When you are done: You have found at least one quarter's retention rate data for the chain and know where it stands in the industry distribution.
Step 2: Compare the direction of TVL and transaction volume changes
What to do: Pull up the chain's TVL and daily transaction volume trends for the past 3–6 months.
How to do it:
Case A (TVL rising, transaction volume also rising): Relatively healthy; users and capital are growing in tandem.
Case B (TVL rising, transaction volume declining): Capital is entering but no one is using it; users are leaving. This is an early warning of a retention problem.
Case C (TVL declining, transaction volume also declining): Double contraction; the chain has entered a decline channel.
When you are done: You have determined which of the above states the chain's "TVL vs. transaction volume" falls into.
Step 3: Look at the trend in active addresses
What to do: Examine the chain's "daily active addresses" curve.
How to do it: 2025 data shows that Base's daily active addresses dropped from 3 million at the start of the year to 452,000 at year-end, an 84.9% decline; Solana fell from 5.1 million to 3.2 million, a 37% decline. These numbers tell you more about whether users are truly doing things on-chain than TVL changes do.
When you are done: You have confirmed whether the chain's active address count is growing, stagnating, or declining.
5. What Kind of New Chain Is Better at Retaining People?
From the data of 2025–2026, several common traits can be summarized:
Deep ecosystem integration: Unichain, built on the Uniswap ecosystem, brought a large amount of genuine DeFi migration and trading activity.
Unique value proposition: Worldcoin (WorldChain) achieved an explosive 1,276% TVL growth thanks to its identity verification system, showing that identity-focused blockchains can form network effects different from purely financial chains.
Real application scenarios, not pure speculation: Ronin's retention rate (19.1%) ranked in the top three, indicating that the repeated engagement mechanisms of a gaming ecosystem are stickier than pure speculation.
Risk reminder: Many projects on new chains are "airdrop-driven." Project parties may use their own funds for incentives to inflate TVL and user numbers in order to secure funding. If the top 10 applications on a chain are all "temporary data farmers," user retention data will quickly expose the real problems. It is advisable to observe a new chain's data for at least 3 months after mainnet launch before deciding whether to commit capital.
After completing the checks above, how do you confirm you truly understand?
Pick a new chain you are watching (e.g., Berachain, Unichain, Ink), open Dune or Artemis, and overlay its TVL, daily transaction volume, and daily active addresses charts. If TVL has dropped more than 50% from its peak while transaction volume and active addresses are declining in tandem, this is not a short-term fluctuation—it is typical "user loss after incentive tapering." If you can clearly explain in your own words why "high TVL but no users" is riskier than "low TVL but high activity," then you truly understand the meaning of retention.
