Band Protocol: Decentralized Data Governance

Authors Soravis Srinawakoon and Sorawit Suriyakarn
Paper Decentralized Data Governance
Version and date Public Draft v3.0.1, May 1, 2019
Design covered Ethereum-era Band Protocol v1.0

This is an educational summary of the 2019 public draft, not a verbatim reproduction or investment advice. The paper’s authors and rights holders retain their rights. The accessible PDF copy is hosted by CoinSwitch.

Original paper Original paper
Directories Ethereum Whitepapers & Research

Decentralized Data Governance is Band Protocol’s May 2019 public draft for a community-governed data layer. It proposes separate governance groups for separate datasets, with dataset tokens used to curate information and a bonding curve linked to the BAND token. The paper’s design is an early Ethereum-based model. It is not the later BandChain architecture that launched on Cosmos SDK.

That distinction matters: later Band materials describe a new chain, oracle scripts, validators, and interchain delivery. Those are later developments, not features to attribute to this 2019 paper.

The problem the paper addresses

Smart contracts cannot independently verify information from outside their blockchain. A decentralized application that needs prices, event results, or other external facts must rely on some data pipeline. The paper argues that a single provider creates a trust and availability risk, and proposes an open framework where communities can curate data feeds and make them available to applications.

Dataset groups, tokens, and BAND

The proposed protocol is organized into dataset governance groups. Each group focuses on a dataset and has its own ERC-20 dataset token. Token holders use that group’s token to take part in curation and governance. In the paper’s design, the group token is backed by BAND through a bonding curve: buying group tokens deposits BAND and mints tokens, while selling them returns a calculated amount of BAND and burns the group tokens. The curve sets the conversion relationship; it does not guarantee that a dataset is accurate or that its token will rise in market value.

The group token and BAND have different roles in this model. The group token provides voting and curation weight for a particular dataset. BAND is the protocol’s native token and the collateral described for issuing those group tokens. The proposal also discusses query fees that can flow to data providers and participating token holders; the fee policy can vary by group and use case.

Two curation models

Token-Curated DataSources

For data that needs ongoing updates, the paper proposes token holders staking dataset tokens behind provider candidates. The candidates with enough backing become active providers. A query can be sent to active providers, and the group’s chosen aggregation logic combines their reports. The paper describes query-fee sharing between providers and the token holders backing them. It does not make one aggregation method mandatory; examples include a median or majority rule.

Token-Curated Registry

A Token-Curated Registry is a public list whose entries can be proposed and challenged. Applicants stake tokens to submit an entry; a challenge can trigger token-holder voting, with deposits and rewards depending on the result. The paper imagines uses such as curated project lists and other public datasets. This is a proposed incentive and governance mechanism, not proof that every listed claim is true.

Risks and limits acknowledged in the paper

The draft itself discusses parasitic services that could copy and resell data, attacks on data providers, collusion, and the difficulty of on-chain voting. It notes that token-based governance can be vulnerable to bribery and participation problems. It also presents slashing as a possible future deterrent, not as an established feature of the design described in this version. A decentralized curation process still depends on the quality, independence, and availability of the underlying sources.

How Band’s design later changed

Band’s own 2019 recap says the team published its whitepaper that year, launched its first mainnet in September, and began developing a new BandChain design later in 2019. The 2020 recap distinguishes that early community-token and bonding-curve model from Band Protocol 2.0, known as BandChain, whose mainnet launched on October 15, 2020. Band announced the Laozi Phase 2 mainnet upgrade in September 2021, adding features including IBC integrations and streamed revenue for data providers.

Current Band documentation describes BandChain data scripts: a script defines which data sources to request and how validators’ reports are combined. Band later introduced its Oracle v3 system, and in 2025 announced that the brand would be called Band. These changes are useful context, but should not be folded into the claims of the 2019 whitepaper.

Related Terms

See Also

Sources