Utility Token Development
Utility token differs from governance token or security token not legally, but functionally: it's needed for something concrete inside protocol. Paying for gas (ETH), computing payment (FIL in Filecoin), service access (API credits), fee discounts (BNB on Binance) — all are utility. Problem with most "utility" tokens: utility is artificial, token isn't needed for protocol to work, forcefully inserted into tokenomics for sale.
Real utility token solves problem that can't be solved without token: coordinated incentives for network participants, trustless escrow, programmable access conditions.
Designing utility mechanics
Token necessity: test
Before designing token, answer: can you replace token with USDC or ETH? If yes — maybe token isn't needed. If no — why not?
Convincing reasons to have own token:
- Governance: token = voting right, ETH can't replace without centralization
- Staking for security: validators stake token, slashing on fraud — skin in game can't replace external asset
- Protocol revenue sharing: tokenholders get part of protocol fees
- Inflationary rewards: bootstrapping subsidy through native token inflation
Capture mechanics
Utility token should "capture" part of protocol value. Popular patterns:
Fee switch: protocol takes X% of operations. Part to treasury, part to tokenholders or buyback/burn. Uniswap governance votes for fee switch exactly this logic.
Staking for access: service providers must stake token. Stake = guarantee of good faith. On violation — slashing. Chainlink operators stake LINK.
Token-denominated pricing: service costs N tokens, not N dollars. Service demand → token demand. Filecoin: storage costs FIL.
Implementation: staking utility
Typical utility token with staking for service access:
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.24;
import "@openzeppelin/contracts/token/ERC20/ERC20.sol";
import "@openzeppelin/contracts/token/ERC20/extensions/ERC20Permit.sol";
import "@openzeppelin/contracts/access/AccessControl.sol";
import "@openzeppelin/contracts/utils/ReentrancyGuard.sol";
contract UtilityToken is ERC20, ERC20Permit, AccessControl, ReentrancyGuard {
bytes32 public constant MINTER_ROLE = keccak256("MINTER_ROLE");
bytes32 public constant SERVICE_ROLE = keccak256("SERVICE_ROLE");
uint256 public constant PROVIDER_STAKE_REQUIRED = 10_000 * 10**18; // 10k tokens
struct ProviderStake {
uint256 amount;
uint256 stakedAt;
bool active;
}
mapping(address => ProviderStake) public providerStakes;
mapping(address => uint256) public serviceCredits; // paid credits
uint256 public constant CREDIT_PRICE = 1 * 10**18; // 1 token = 1 credit
uint256 public burned;
event ProviderRegistered(address indexed provider, uint256 stake);
event ProviderSlashed(address indexed provider, uint256 amount, string reason);
event CreditsPurchased(address indexed user, uint256 amount);
constructor(address admin, address treasury, uint256 initialSupply)
ERC20("Utility Token", "UTL")
ERC20Permit("Utility Token")
{
_grantRole(DEFAULT_ADMIN_ROLE, admin);
_grantRole(MINTER_ROLE, admin);
_mint(treasury, initialSupply);
}
// Provider stakes tokens to register
function stakeAsProvider() external nonReentrant {
require(!providerStakes[msg.sender].active, "Already registered");
require(
balanceOf(msg.sender) >= PROVIDER_STAKE_REQUIRED,
"Insufficient balance"
);
_transfer(msg.sender, address(this), PROVIDER_STAKE_REQUIRED);
providerStakes[msg.sender] = ProviderStake({
amount: PROVIDER_STAKE_REQUIRED,
stakedAt: block.timestamp,
active: true
});
_grantRole(SERVICE_ROLE, msg.sender);
emit ProviderRegistered(msg.sender, PROVIDER_STAKE_REQUIRED);
}
// User buys credits
function purchaseCredits(uint256 creditAmount) external nonReentrant {
uint256 tokenCost = creditAmount * CREDIT_PRICE;
require(balanceOf(msg.sender) >= tokenCost, "Insufficient tokens");
// 80% burn, 20% to treasury — deflationary mechanic
uint256 burnAmount = tokenCost * 80 / 100;
uint256 treasuryAmount = tokenCost - burnAmount;
_burn(msg.sender, burnAmount);
burned += burnAmount;
_transfer(msg.sender, treasury, treasuryAmount);
serviceCredits[msg.sender] += creditAmount;
emit CreditsPurchased(msg.sender, creditAmount);
}
// Provider consumes credits for service
function consumeCredits(address user, uint256 amount) external onlyRole(SERVICE_ROLE) {
require(serviceCredits[user] >= amount, "Insufficient credits");
serviceCredits[user] -= amount;
}
// Slashing on provider violation
function slashProvider(
address provider,
uint256 amount,
string calldata reason
) external onlyRole(DEFAULT_ADMIN_ROLE) {
ProviderStake storage stake = providerStakes[provider];
require(stake.active, "Not active provider");
require(amount <= stake.amount, "Exceeds stake");
stake.amount -= amount;
_burn(address(this), amount); // burn slashed tokens
burned += amount;
if (stake.amount < PROVIDER_STAKE_REQUIRED / 2) {
stake.active = false;
_revokeRole(SERVICE_ROLE, provider);
}
emit ProviderSlashed(provider, amount, reason);
}
}
Unstaking cooldown
Provider shouldn't withdraw stake instantly. Cooldown period protects from slash evasion (discovered problem → quickly withdrew stake → no slashing):
uint256 public constant UNSTAKE_COOLDOWN = 14 days;
mapping(address => uint256) public unstakeRequestedAt;
function requestUnstake() external {
require(providerStakes[msg.sender].active, "Not active");
unstakeRequestedAt[msg.sender] = block.timestamp;
providerStakes[msg.sender].active = false;
_revokeRole(SERVICE_ROLE, msg.sender);
}
function finalizeUnstake() external nonReentrant {
require(unstakeRequestedAt[msg.sender] > 0, "No unstake request");
require(
block.timestamp >= unstakeRequestedAt[msg.sender] + UNSTAKE_COOLDOWN,
"Cooldown not elapsed"
);
uint256 amount = providerStakes[msg.sender].amount;
providerStakes[msg.sender].amount = 0;
unstakeRequestedAt[msg.sender] = 0;
_transfer(address(this), msg.sender, amount);
}
Supply distribution
Typical distribution for protocol utility token:
| Allocation | % | Vesting |
|---|---|---|
| Team and advisors | 15–20% | 4 years, cliff 1 year |
| Investors | 15–25% | 2–3 years, cliff 6 months |
| Ecosystem/grants | 20–30% | Linear 3–5 years |
| Liquidity/DEX | 5–10% | TGE or as needed |
| Treasury | 20–30% | Governance decides |
| Public sale / IDO | 5–15% | Partial TGE |
Total TGE (Token Generation Event) — ideally not more than 15–20% supply. Too large TGE float creates selling pressure.
Avoiding anti-patterns
Useless buyback: buyback token from treasury and burn — this is value transfer from treasury to tokenholders. Doesn't create value itself. Makes sense only if protocol has real revenue.
Circular dependency: token needed to use protocol, protocol needed to get token. Without external value — closed loop.
Governance without power: governance token with no real right to change protocol parameters — decorative. Users understand this.
Timeline for utility token with basic staking mechanics and credits system: 1–2 weeks including tests. Complex tokenomics with multiple mechanics — 3–4 weeks.







