This article was originally published at https://mirror.xyz/lighthousegov.eth I invite readers to subscribe and join our growing community of practitioners and enthusiast's https://docs.lighthouse.cx/community or feel free to reach out directly on my Telegram.
Signals is an on-chain protocol that converts latent community sentiment into a real-time, ranked idea board of DAO initiatives. Inspired by Conviction voting, Commitment voting and veToken Systems like Curve, Aragon and Velodrome, it lets members lock their existing governance tokens for a period of time to signal their support for what they think is most important. This token-time locking, coupled with configurable decay curves and acceptance thresholds, equips decentralized organizations with a transparent, mathematically-grounded mechanism for continuous sensemaking.
We envision this protocol becoming an invaluable tool for on-chain organizations to gain clarity on structural changes and where to allocate capital. Whilst also unlocking a new set of utility for their underlying governance token.
After a decade of experimenting with on-chain governance, most DAOs still struggle to turn community intent into timely, accountable action. Token-based voting commonly employed by DAOs can capture a binary "yes or no" on formalized proposals, but it does little to surface preference intensity of community participants. This makes it a poor tool for uncovering what should be worked on next or how strongly members care about different initiatives.
Many DAOs have tried to solve these shortcomings through structure and process, but this has resulted in many familiar frustrations:
Endless Discord and forum threads with no clear next steps
Committees that drift or concentrate power
Fragmented knowledge bases that deter newcomers
Social pressure that favors incumbents over fresh ideas
Although today’s largest DAOs commonly make use of Web3 tools for trustless voting and treasury management, for everything else they are relying on traditional systems which bring with them the bureaucracy of yesterday—and it is hindering their ability to run efficiently.
By leveraging community members’ preference intensity as expressed through time-lock commitments, Signals constructs a dynamic, real-time idea board where the most widely-held opinions rise naturally to the top.
Instead of proposals, Signals deals with “Initiatives” — an idea, course of action, or other concept that is seeking support from the community. Participants submit Initiatives to a Signals space (called a “Board”) which has been configured by the DAO and usually revolves around a specific theme. A DAO can have multiple Boards, all configured differently and for different purposes. Board configurations are all on-chain, so they could be configured trustlessly as the result of a DAO vote or existing trust models.
Users lock up their existing tokens in support of a specific Initiative and for a specified duration, which demonstrates a genuine commitment through opportunity cost. Voting against an Initiative is not possible. By eliminating low-conviction indicators, every signal becomes an important one.
Boards will typically be configured to use DAO governance tokens, though any asset (token or stablecoin) could be specified. Each lockup is minted as a transferrable NFT bond which allows the holder to redeem the underlying tokens once the lock expires, opening the door for secondary markets and other qualitative measurements.
Signals handles time as “Epochs,” a configurable base unit of time that represents the level of urgency most appropriate for a particular Board. Users who choose to lock their tokens for more than one Epoch contribute additional support to their chosen Initiative. The formula for calculating support can be customized as needed, but a standard starting point could look like the following, which we feel would provide enough flexibility for the majority of use cases:
S=T×Dk
Where:S=The starting amount of support providedT=The number of tokens locked upD=The lock duration, denoted as a number of Epochsk=Exponential bonus for longer lockups
Calculating support in this manner, and displaying the details, provides several important properties:
Users with fewer tokens can compete with larger holders, who are less likely to commit to a longer lock duration
Communities can compare the levels of conviction shown by small vs large holders
Support measurements become more resilient to manipulation
Time becomes a democratizing factor in governance
The amount of support provided by an individual lockup decays over time, to ensure that indicators of support remain current and relevant. Each Board can configure the decay curve formula that is used to calculate each Initiative’s support value at the current time. Decay curves could be linear, exponential, or something much more complicated.
Calculating the total amount of current support behind an Initiative is expressed as the sum of support provided by all lockups, given the current time:
CurrentSupporti(t)=∑ℓ∈LocksiWeight(ℓ,t)
Where:CurrentSupporti(t)=An Initiative’s support at time ti=The Initiative we are assessingLocksi=The set of locks supporting this Initiativel=One specific lock in that setWeight(ℓ,t)=Individual value of support a lockup provides at time t
Using decay introduces a powerful governance dynamic:
Initiatives must maintain ongoing support to remain relevant
Abandoned Initiatives naturally lose prominence
Early support provides a head start, but not permanent advantage
Decay creates a cost to waiting, which aligns contributors around timely action
Users are further dissuaded from supporting unpopular Initiatives
Boards are configured with a target level of support, called the “acceptance threshold”. Any Initiative that currently has enough support to exceed this threshold can be “actioned,” in which case the Initiative is closed and all tokens locked up in support of that Initiative are immediately unlocked. This early unlock incentivizes users to only support Initiatives they think will be popular. It is up to DAOs to decide how to treat initiatives that have been actioned.
Actioning could be configured to happen automatically once the criteria is met, or it could require manual activation by a committee, DAO vote, or other mechanism.
Below you will find some simulations of the manner in which support can decay over time, with each color representing a contribution from a different supporter.
For DAOs that support delegating voting power, delegates would not be able to fully participate in Signals because they do not directly own the tokens that are entrusted to them. To account for this, a Board can be configured to make use of “Endorsements,” a binary signal of support that can be issued by a user based on their amount of delegated voting power.
The Board could require all Initiatives to receive a certain number of Endorsements in addition to locked up support before they can be actioned. There could also be a maximum limit to the number of Endorsements each Initiative can receive, which encourages delegates to act early, to avoid missing out on the reputation gained from publicly supporting a popular Initiative.
Conviction Voting lets holders stake without fixed locks; conviction builds the longer tokens stay committed and can be re-allocated at any time (which resets the build-up) Proposals are automatically executed—often with treasury funds released—once time-weighted conviction crosses an adaptive threshold tied to the amount requested.  
Conviction Voting translates persistent, growing conviction directly into binding, resource-allocating decisions.
Commitment Voting is an aspect of proposal/election voting in which winners prove their dedication to their choice. A voter commits to locking up a number of tokens for a period of time, which determines the voting power added to the voter’s selected candidate. If the voter’s selection wins the election, their tokens continue to be locked up as committed; voters who supported the opposing side immediately have their committed tokens returned to them.
Commitment Voting channels intensity into a binding choice where victory comes with an opportunity cost and there are few downsides to supporting a losing candidate.
Curve’s veCRV, Aragon OSx gauges, and Velodrome have users lock up tokens for a duration of time as a prerequisite for receiving voting power. Longer lockups provide more voting power, which can be used in all votes for the duration. These systems allow users to convey their intensity of preference for participating in voting in general, but does not convey intensity of preference on specific votes.
The Signals proof of concept was developed during the Arbitrum Collabtech Hackathon, where it won first place. We continued to work on our Bond Marketplace (a Signals sub-system) during the Uniswap Hooks Incubator which won prizes from Arbitrum and the Uniswap Foundation.
All smart contracts and web app dashboard code is open source and can be found at https://github.com/0xLighthouse/signals
We are now seeking the brightest minds to work with us on producing a whitepaper to formally prove the most interesting properties of the protocol:
Capturing voter preference intensity
Opportunity cost as sufficient risk
Locking mechanisms improving sybil resistance
Gamifying Endorsements to trustlessly reward and build reputation
Empowering smaller voting blocks increases inclusivity
These collective properties are poised to solve some of Ethereum’s biggest challenges when it comes to on-chain sensemaking and capital allocation.
If you are interested in Signals please reach out.
Conviction voting, https://github.com/1Hive/conviction-voting-cadcad/blob/master/README.md
Curve Finance. 2025. Understanding veCRV. [online] Curve Resources. Available at: https://resources.curve.fi/reward-gauges/understanding-vecrv
DAOstar. 2025. DAO Delegate Incentive Programs: How to build for the future. [online] Available at: https://daostar.org/reports/delegate.pdf
Curve DAO (n.d.) ‘Locked CRV (veCRV) – Overview’, Curve Resources. Available at: https://resources.curve.fi/vecrv/overview/
Berg el al. (2020) Commitment Voting: A Mechanism for Intensity of Preference Revelation and Long-Term Commitment in Blockchain Governance . http://dx.doi.org/10.2139/ssrn.3742435
Emmett, J. (2019) ‘Conviction Voting: A Novel Continuous Decision Making Alternative to Governance’, Medium (Giveth blog), 3 July. Available at: https://medium.com/giveth/conviction-voting-a-novel-continuous-decision-making-alternative-to-governance-aa746cfb9475
Khezr et al. (2024) ‘Voting with Time Commitment for Decentralized Governance: Bond Voting as a Sybil-Resistant Mechanism’, Management Science (forthcoming). SSRN pre-print, doi: 10.2139/ssrn.4259599. Available at: https://ssrn.com/abstract=4259599
Aragon (2024) ‘Introducing Mode Governance’, Aragon Blog, 16 October. Available at: https://blog.aragon.org/introducing-mode-governance/
Velodrome (2022) ‘Velodrome: Launch. Part 1 – Overview’, Medium, 19 May. Available at: https://medium.com/@VelodromeFi/velodrome-launch-part-1-25b38120224a
