Demand sinks and useful token economies
A token economy needs more than managed supply. It needs recurring use that rewards the people and processes making the network useful.
On this page
Start with demand
Issuing a token before establishing demand creates a structural problem: supply has a schedule, while the reason to use it remains unproven.
The team obsesses over vesting. The community watches unlocks. The chart becomes a morality play about emissions, market makers, and timing. Everyone talks about supply pressure because supply pressure is visible.
Demand needs a harder test.
A disciplined issuance schedule does not establish lasting demand. The network needs repeated, credible reasons to acquire, spend, lock, or hold its token in order to do useful work. A compulsory payment step is not enough. The underlying service has to justify its cost.
Interest is not recurring use.
Speculation is not recurring use.
Community excitement is not recurring use.
Belief in the project is not recurring use.
The demand this essay is concerned with comes from use: people or agents acquire the token because it has a defined role in a service they choose to use. That is distinct from buying it in anticipation of a price change.
Narrative Token
Demand Sink Token
The question is not whether a token moves. The question is whether it returns with more reason to exist than before.
What a demand sink actually is
A demand sink ties token use to a concrete network function. Payment transfers tokens. Collateral locks them. Burning removes them. These mechanisms affect circulation differently; none establishes recurring demand for the service by itself.
The categories are specific.
Payment for execution.
Payment for access.
Payment for prioritisation.
Payment for licensing.
Payment for routing.
Collateral for truth telling.
Stake for curation.
Bond for reputation.
Reservation of scarce capacity.
Coordination rights that cannot be obtained any other way.
Whatever form it takes, the token’s role must be tied to the work the network does. A requirement creates a payment route. A useful service gives people a reason to take it. The economic argument needs both.
The aim is structural pull. Not excitement alone. Demand tied to use. It lasts only while the network continues to deliver something worth its cost.
Recurring demand starts with useful work. The payment design determines how that demand reaches the token.
Supply leakage and productive use
Supply is not the enemy. It has to be considered alongside demand, circulation, and the incentives given to recipients.
Supply leakage describes value leaving without reinforcing useful activity. The sell-side flow matters: who receives tokens, what work they perform, and how much they sell. Selling is not itself a failure; contributors have costs. The question is what the distributions fund and whether that work earns further use.
Value can leave the network while useful work continues. The question is whether distributions support that work or merely sustain participation for the next payout.
When distributions do not fund useful work or retained participation, the network is supporting a distribution schedule rather than strengthening its service.
Locking tokens to earn more tokens can defer circulating supply without creating demand for the underlying service. Staking has a stronger economic purpose when it secures work, backs commitments, or exposes participants to a defined risk. The test is the function it supports, not the lock-up alone.
Subsidies can fund early participation. They become a substitute for demand when activity has no reason to continue without them. The test is what remains after the incentive: useful infrastructure, better service, and participants who return for the work itself. Rented attention is not the same as retained use.
Where does demand come from.
What exactly requires the token.
Who receives it once spent.
What percentage leaks out immediately.
What percentage returns to the people
who strengthen the network.
If you cannot map those flows precisely, you do not have token design. You have issuance plus hope.
Percentages on a launch chart do not tell you whether value reinforces the network or drains out of it.
Internal demand flywheels
A serious token system should be able to answer this clearly: when someone spends the token, who gets stronger?
A treasury can fund useful infrastructure. The question is whether spend reaches the work that improves the service, and whether that improvement earns continued use. A payment gate alone does not answer it.
The design worth pursuing connects useful spend to productive work. Some receipts cover operating costs. Some fund the treasury. Some reward contributors who improve supply quality, coverage, curation, verification, or execution reliability. Burning is a separate supply choice, not a substitute for those functions.
That redistribution ties usage back to contribution. The intended loop is clear: spend rewards useful work, the work improves the service, and the improved service earns repeat use. Each link has to hold.
We should call it a flywheel when improved service produces retained use, not merely when tokens move through a circular diagram.
The loop we are designing connects redistribution to attributable contribution. Not passive holding. Not liquidity farming. Work that can be identified, assessed, and rewarded over time.
That is where token design connects directly to identity and authority architecture. If you cannot track who contributed what, when, and under what terms, you cannot route value to the right people. You end up rewarding presence instead of performance. And presence-based rewards are just another form of supply distribution pretending to be a flywheel.
Our priority is to make contribution inspectable: attributed actions, curated data, assessed outputs, maintained infrastructure, resolved disputes. Identity and authority establish who acted and under which permissions. Evidence and evaluation are needed to assess what the contribution was worth.
That link is what we want redistribution to reinforce: useful contribution, assessed against the work the network needs.
Select a token design scenario
Explore how payment routing and repeat usage change this model’s reinvestment balance.
Our aim is for useful spend to reward the people who make the next service worth buying.
The model compares allocation choices under chosen assumptions. It does not estimate adoption, token price or investment returns.
When launch incentives outrun use
Projects that create issuance, rewards, and reasons for early participation still have to establish a reason for repeated use. Launch activity does not answer that question.
A subsidised launch can obscure that gap. If participation depends on rewards rather than useful service, reducing the subsidy exposes the underlying demand. The useful measure is what participants continue to do when the launch incentive ends.
They distribute ownership before they create necessity.
They produce supply before they earn demand.
They reward presence before they reward contribution.
They market the token before they hard-wire its role.
Then, when the pressure arrives, they discover they built a chart before they built an economy.
Managing unlocks is part of the design, not the whole design. Our bet is on useful work earning repeat use, with spend reinforcing the contributors who make that work possible. Even then, transaction volume alone does not establish durable token value: circulation, recipient selling, supply, and the balances users need to hold all matter.
Agents as users of useful services
Agents act under the goals, budgets, permissions, and selection rules set by their operators. Access, execution, latency, quality, cost, and reliability give us concrete criteria for evaluating the services they use.
An agent using a token-funded service still needs that service to fit its task and its operator’s constraints. Data, execution, verification, storage, routing, and licensing are concrete functions to evaluate. The token route has to justify its cost against available alternatives; repeated use has to be earned.
That is where the future gets real. But it requires honesty about what AI actually changes.
Putting AI in the deck does not create demand. Saying agents will use the network does not create demand. We are betting on useful, metered services whose quality and cost justify repeated use, and on records that let productive contribution be recognised.
What exactly is the agent buying.
How is that priced.
Why must the token be used.
Where does the spent token go.
Which contributors are rewarded.
How does that improve future service quality.
If you cannot answer those questions, then AI is just another narrative wrapper around weak internals.
Quality
—
Freshness
—
Coverage
—
Reliability
—
AI demand is only useful when it buys something the network must keep earning.
The work we are committing to is making useful activity and productive contribution visible, so economic choices can be tested against what the network delivers.
Useful work first.
A token economy needs more than managed supply. It needs recurring use that rewards the people and processes making the network useful.