A period in which protocols distributed tokens to attract deposits generated enormous headline figures and left a clear record of what such incentives achieve.
The mechanism
Protocols issued tokens to users who deposited assets or provided liquidity.
Which produced advertised returns far above anything sustainable.
The return was denominated in newly issued tokens rather than in fees earned.
The growth
Deposits across protocols rose dramatically over months.
Which was reported as adoption.
Much of it was capital moving between protocols chasing the highest current incentive.
Mercenary capital
Deposits that leave when incentives end.
Which was documented clearly when programmes concluded.
Retention after incentive expiry was low across most protocols studied.
Token price effects
Recipients selling issued tokens created continuous downward pressure.
Which meant advertised returns fell as the token price fell.
The yield was self-defeating in a fairly direct way.
Forks and copies
Protocols were copied with modified incentives within days.
Which was possible because the code was public.
A substantial number of these copies were fraudulent or abandoned.
Rug pulls
Projects removing liquidity or minting tokens and disappearing.
Which was common enough during the period to be a recognised category.
Contract review would have identified the mechanisms in most cases.
What survived
Protocols with genuine fee revenue and users who would have been there without incentives.
Which is a small subset of what existed at the peak.
These are identifiable by revenue that persisted after incentives ended.
The design lesson
Incentives can bootstrap a network and cannot substitute for a reason to use it.
Which is now widely accepted and was contested at the time.
This is historical description rather than advice about any protocol.
Governance token distribution
Incentive programmes distributed governance rights to depositors.
Which was intended to decentralise control.
Recipients selling immediately meant governance concentrated among buyers rather than users.
Total value locked as a metric
Deposits became the headline measure of protocol success.
Which is easily inflated by incentives and by double counting through wrapped assets.
The metric's prominence encouraged exactly the behaviour that made it meaningless.
Composability risks
Strategies stacked across multiple protocols multiplied dependencies.
Which produced cascading failures when one component broke.
Yield aggregators automating these stacks concentrated the exposure.
What replaced it
Fee-sharing models, points programmes and more targeted incentives.
Which are refinements of the same idea with more attention to retention.
Assessing a protocol now
Revenue from actual usage, retention after incentives ended, and what the token is required for.
Impermanent loss awareness
Advertised returns rarely accounted for divergence loss on provided liquidity.
Which meant headline yields overstated actual outcomes substantially.
Tools calculating historical position outcomes were built afterwards and show this clearly.
Smart contract risk concentration
Capital chasing yield concentrated in newly deployed unaudited contracts.
Which produced a high rate of loss from exploits during the period.
Incident frequency correlated with the launch rate of new protocols.
Anonymous teams
Many protocols were launched by unidentified developers.
Which removed accountability entirely when things failed.
Some anonymous teams built durable protocols, which complicates any simple rule.
Regulatory attention afterwards
Questions about whether yield programmes constituted securities offerings.
Which has been addressed in enforcement actions and guidance.
The lasting infrastructure
Automated market makers, lending protocols and composability standards developed during the period and remain in use.
Why it still matters
The period demonstrated conclusively that paying people to use something produces usage that ends when the payment does.
That is not a surprising finding, and enormous sums were deployed on the assumption that it would not hold.
The infrastructure built during the period is the durable output; the deposit figures were not.
Reading the primary material
Deposit data before and after incentive programmes ended is fully public.
Which shows retention directly rather than by inference.
The assessment question
What revenue does this protocol generate from actual usage, and what happened to deposits when the incentives stopped.
What replaced the metric
Protocol revenue, active user counts and retention have partially displaced deposit totals as the headline measure.
Which is a genuine improvement, since revenue is difficult to fabricate and is verifiable on chain.
Deposit figures continue to be quoted in marketing because they remain the largest number available.
The one sentence version
Paid usage stops when the payment does, and the buildings left behind are worth more than the traffic that was rented.
A note on what was learned
The period is easy to dismiss and it produced the automated market maker designs, lending protocols and composability standards that the field still runs on.
The incentive programmes were the wrapper; the infrastructure was the substance, and separating them took several years.
The verifiable test
Look at a protocol's deposits before and after its incentive programme ended.
The chart answers the question in seconds and is public for every protocol from the period.
Most of them look the same, which is the finding.