Incentives buy deposits. They don’t buy loyalty.
Every DeFi team learns this, usually about a month after emissions taper and the chart falls off a cliff. The TVL number describes your incentive budget, not your product, and the gap between the two is the only thing worth measuring.
The question behind the number
Why is the value there?
If the answer is because we’re paying for it, you know what happens when you stop. That’s not necessarily a mistake — bootstrapping a market requires liquidity to exist before anyone will use it, and incentives are the standard way to create that.
The mistake is mistaking rented liquidity for product-market fit and building plans on top of deposits that were always leaving.
What actually retains liquidity
Better execution. If routing through your protocol genuinely gets a better price, integrators route through it. That’s structural and it survives emissions ending.
Lower risk at comparable yield. Depositors are more risk-sensitive than the yield-chasing narrative suggests, particularly the larger ones. A protocol with a cleaner security record and clearer risk documentation keeps deposits at lower headline rates.
Integration dependency. When another protocol, a wallet or an application depends on you, that liquidity is stickier than any incentive. Integrations are the most underrated growth channel in DeFi.
Real fee revenue. Yield generated by actual usage rather than emissions is qualitatively different, and sophisticated depositors distinguish between the two immediately.
Something a depositor can’t get elsewhere. A market that doesn’t exist on other venues, an asset nobody else supports, a risk profile that’s genuinely distinct.
Structuring incentives properly
Define what you’re buying. Depth in a specific pair. Usage of a new market. An integration with a named partner. General TVL growth is not a goal, it’s a number.
Taper from the start, and publish the schedule. Depositors plan around it, which is better for everyone than a surprise reduction.
Decide in advance what the number should be when emissions stop. That figure is your actual product. Everything above it is rent.
Watch concentration. If a handful of addresses hold most of the deposits, your TVL is one decision away from halving. Depositor count and distribution matter as much as the total.
Measure fee revenue separately from incentive-driven activity. Wash-like behaviour to farm rewards inflates volume and tells you nothing.
Trust is the growth channel
In DeFi, security posture is marketing whether you treat it that way or not. The things that move deposits from sophisticated users:
- Published audits, in full, including findings you chose not to fix and the reasoning.
- A documented risk model. What happens in the tail cases, stated plainly.
- Oracle dependencies disclosed. Which feeds, what happens if one fails, what the fallback is.
- A clear position on admin keys. Who can upgrade, what the timelock is, who holds the multisig. Evasiveness here is read correctly as a warning.
- Incident communication. Speed and specificity, and the second update landing when you said it would.
Users in this category are unusually sophisticated about risk and unusually unforgiving about concealment. The protocols that recover from incidents are the ones that communicated well during them.
The growth channels that work
Integrations. Being the venue another protocol routes through, or the market a wallet surfaces. Business development rather than marketing, and the highest-return activity available.
Technical content. Explaining your mechanism properly, publishing analysis of your own data, writing the piece that finally clarifies a hard concept. This reaches exactly the people who allocate capital.
Developer accessibility. Documentation, SDKs and examples that let someone integrate without asking you questions.
Governance participation. An active, informed governance process is a retention mechanism and a signal that the protocol is genuinely decentralised rather than describing itself that way.
Selective creator work, aimed at educated audiences. DeFi audiences discount promotional content heavily, so this only works with creators who actually explain things.
What drives the work
- Complexity of the mechanism, since harder-to-explain protocols need far more educational content.
- Number of chains deployed, each with its own community and integrations.
- Security posture, and how much documentation and audit work already exists.
- Whether incentives are running, which changes both the measurement and the message.
- Integration pipeline, since business development effort scales with how many partners you’re pursuing.
- Governance maturity, if a live DAO needs supporting.
The summary
Publish the number that survives emissions ending. Internally at minimum, publicly if you’re confident.
Then spend on the things that make that number bigger — integrations, execution quality, security posture, documentation — rather than on the incentives that inflate the other one.
Rented liquidity is a legitimate bootstrap and a terrible strategy. The transition between the two is where most DeFi protocols either become durable or quietly stop mattering.
Common questions
Is total value locked a useful metric?
Only alongside the question of why the value is there. TVL bought with emissions leaves when emissions stop, which means the number describes your incentive budget rather than your product. The more useful measures are what share of deposits remain after incentives taper, how concentrated deposits are among a few addresses, and whether the protocol generates fee revenue independent of rewards.
How do you keep liquidity after incentives end?
By giving depositors a reason to stay that is not the reward — better execution, lower risk, an integration they depend on, or yield generated by real usage rather than emissions. Protocols that retain liquidity generally offer something structurally useful. Protocols that lose it were paying rent and the tenancy ended on schedule.
What is mercenary capital and does it matter?
Mercenary capital is liquidity that moves to wherever the incentive is highest, with no attachment to the protocol. It is not inherently bad — it can bootstrap a market and make early usage possible. It becomes a problem when a team mistakes it for product-market fit and builds plans around deposits that were always going to leave.
How should a DeFi protocol structure incentives?
With a defined purpose, a taper, and a measure of what you are buying. Incentives work as a bootstrap for a specific goal — depth in a particular pair, usage of a new market, integration with a partner. They fail as a permanent operating expense. Decide before launch what the number should look like once emissions stop, because that is the only figure that describes the actual product.
What matters most for DeFi user trust?
Security posture and transparency, in that order. Published audits including unfixed findings, clear documentation of the risk model, disclosed oracle dependencies, a stated position on admin keys and upgradeability, and honest communication when something goes wrong. Users in this category are unusually sophisticated about risk and unusually unforgiving about concealment.
