Key takeaways

• Liquidity mining pays depositors in a protocol's own token, and it reliably buys launch TVL (total value locked, the dollars users have deposited) but rarely keeps it once the rewards stop.

• Berachain pulled in $3.1 billion of pre-deposits before mainnet and now shows about $36 million in TVL on DefiLlama, roughly 99% below its peak.

• The honest test of any incentive program is how much capital survives the taper, and one benchmark puts organic growth at 40% or more of peak TVL retained against below 20% for rented growth.

Liquidity mining is the fastest way to put a big number on a new DeFi protocol, and one of the fastest ways to lose it again. Every yield product launches with a screenshot of a large TVL figure, a green line and a pre-deposit campaign that filled up in hours. The founder posts it, the community cheers, and the number is the story for about a week, until the emissions taper and the capital walks out the door.

Berachain, the layer-one chain built around rewarding liquidity providers, is the cleanest recent case. Its pre-deposit campaigns pulled in $3.1 billion before mainnet went live, and DefiLlama's data now puts the chain's TVL at about $36 million, roughly 99% below where it peaked in early 2025. Uniswap ran the same experiment five years earlier, and when its UNI rewards ended its liquidity fell about 40% in a single day.

When a protocol asks us to grow TVL, we split the brief into the two questions it usually hides, which are how you get capital in and how you make it stay. The first is mostly a spending problem, because anyone with a big enough incentive budget can rent liquidity for a launch. The second is a product and marketing problem, and it is the only one that compounds.

A launch high tells you what you were willing to spend. Whether any of it stays is a separate question, and it is the only one worth reporting.

How does liquidity mining work?

Liquidity mining works by paying people in a protocol's own token for depositing assets into its pools. Users lock capital, such as a pair of tokens on a decentralized exchange or stablecoins in a lending market, and earn newly issued tokens on top of any trading fees, usually in proportion to their share of the pool and only for as long as they stay.

Protocols do it because an empty pool is a broken product. Traders need deep pools to swap without moving the price, and borrowers need lenders before a lending market can function, so a new protocol has to attract capital before it has any users to show. One DeFi analytics team calls liquidity mining one of the most wildly successful ways to get participants to come to a protocol, and on day one that is true.

The catch sits in the same analysis, which describes bootstrapped liquidity as almost entirely mercenary, there for the yield rather than the product. Uniswap showed what that means in November 2020. It had been paying 2.5 million UNI a month to providers in four pools, and when the program ended its TVL dropped from $3.07 billion to $1.75 billion almost overnight.

What is the difference between liquidity mining and yield farming?

Liquidity mining is one specific way to earn yield, where a protocol pays you its own token for supplying liquidity to its pools. Yield farming is the wider habit of moving capital between protocols to chase the best available return, and those token rewards are usually what yield farmers are chasing, so every program you launch gets compared against every other one.

That comparison is why mercenary liquidity exists, and the label isn't really an insult. It is a rational response to an offer. When the only reason to be somewhere is the payout, the exit is priced in from the start, and a farmer with a spreadsheet will move the moment a rival protocol pays a point more.

Berachain tried to design around this. Its proof-of-liquidity model tied network security to DeFi participation so that liquidity would be "sticky," yet TVL still fell 88% to about $393 million within a year as the yields that attracted it dried up. Clever mechanism design didn't change why the capital came, so it didn't change why the capital left. Crypto ran the same experiment on people rather than dollars with play-to-earn games that paid players wages, and got the same answer.

How much liquidity should survive when the rewards stop?

A healthy liquidity mining program keeps a meaningful share of its peak after the incentive tapers. One practitioner benchmark draws the line at 40% or more of peak TVL retained for organic growth, against a fall below 20% within two weeks for the artificial kind. The two curves look identical on launch day and nothing alike ninety days later, which is inconvenient, because ninety days later is where the real business lives.

Users behave the same way as capital. The same analysis notes that even in normal conditions only about 25% of first-time DeFi users become regular users. During an incentive campaign that figure gets worse and harder to see, because the reward distorts why anyone showed up in the first place, and you can't read intent off a number you paid to manufacture.

This is why the sharper teams we sit across from have started asking for something different. Instead of one more trading competition to spike volume for a week, they want sticky TVL and holders who are still around next quarter. It is a harder brief to deliver, and it is the one worth taking, because it is also the work of crypto user retention that most launches skip entirely.

What we'd tell a founder about liquidity mining

Give capital a reason to stay that isn't the yield. Incentives win the first deposit and utility wins the second month. For a yield-bearing asset that means integrations where the token is genuinely useful, as collateral, in payments or as a building block inside someone else's product, so that leaving costs something beyond a forgone reward. We treat those integrations as launch infrastructure rather than a phase-two extra.

Reward the actions that predict staying. Paying people to park capital rewards parking. Paying for the behaviour that correlates with retention, like durable liquidity, using the asset inside the protocol and referring users who also stick around, buys something you want to keep. The emissions cost the same either way, and the difference is what they leave behind once they stop.

Separate the cohort you rented from the cohort you earned. From day one we tag inflows by source and track each group's decay curve on its own. Incentive-driven capital and organic capital behave nothing alike, and blending them into one TVL line hides the signal that predicts what happens after the campaign.

Put retention in front of the founder, not the peak. A launch-day high is a nice screenshot and a poor forecast. Week-four and week-twelve holder retention, net of incentive outflows, is what we report, because it is the figure an investor should underwrite and the one that predicts the next raise.

Incentives are the opening move

None of this means incentives are bad. It works, and our team runs pre-deposit pushes, points programs and competitions, because they are legitimate tools. The mistake is treating the number they produce as the finish line. The stakes are rising too, because yield-bearing stablecoins alone grew about 300% in a year to roughly $22 billion by November 2025, and all of that capital shops for the best return and never stops shopping. If your whole go-to-market is a bigger offer than the last protocol's, you are one epoch away from being someone else's outflow, so before you post the launch screenshot, make sure you can answer who still has a reason to be here on the morning the rewards stop.

Frequently asked questions

What is liquidity mining in crypto?

It is a DeFi incentive where a protocol rewards users with its own tokens for depositing assets into its pools. It is how new exchanges and lending markets attract capital before they have organic users. Uniswap's 2020 program paid 2.5 million UNI a month across four pools, and liquidity fell about 40% in a day when it ended.

What is mercenary liquidity?

Mercenary liquidity is capital that arrives for incentive rewards and leaves as soon as a better offer appears or the rewards stop. It is rational behaviour rather than bad faith. Berachain's TVL peaked at $3.35 billion in 2025 after large pre-deposit campaigns and fell 88% within a year as yields dropped, which is the pattern in its clearest form.

Why does TVL drop after an incentive program ends?

TVL drops because most of the capital came for the reward rather than the product, so the reason to stay disappears with the emissions. One benchmark treats a fall below 20% of peak within two weeks of the taper as the mark of artificial growth, while organic programs hold 40% or more of their peak.

How do you measure whether liquidity mining worked?

Measure what remains after the rewards taper, not the launch peak. Track retained TVL as a share of peak, split capital by whether it arrived through incentives or organically, and follow holder retention at week four and week twelve. Only about 25% of first-time DeFi users become regular users even without incentives, so that is a realistic baseline to beat.

Is liquidity mining still worth running in 2026?

Yes, as a launch tool, but not as a growth strategy on its own. Competition for yield is fiercer than ever, with yield-bearing stablecoins up about 300% in a year to roughly $22 billion by late 2025. A program works when it is paired with real utility and rewards for behaviour that predicts retention, so capital has a reason to stay when emissions end.