Quick note: seems like DMs are disabled here. If you’d like the full Governance Audit report, just drop me a note at [email protected] and I’ll send it over.
Update: Delegate Vote Quality — Lido at 1.8 median
Following up on my earlier governance health analysis — we’ve now shipped Delegate Vote Quality scoring.
Four signals per delegate: Deliberation, Independence, Focus, and Originality. Across 24 DAOs, 75%+ of delegates score below 5/10.
Lido’s median VQS is 1.8 — the lowest in our dataset. The pattern: rapid voting with high consensus. Most delegates vote the same way within very short timeframes.
This aligns with what BCV and others have raised about concentration — it’s not just about who holds voting power, but about how that power is exercised.
Full analysis: chainsights.one/blog/delegate-vote-quality-75-percent
Lido delegate scores: chainsights.one/matrix/lido
The question for Lido governance: does rapid consensus reflect efficient decision-making, or a lack of independent deliberation?
2 Likes
Lido DAO lacks a decentralized oversight and ownership layer. In its current form, delegates resemble a small group of paid advisors to Lido Labs. It’s not aligned with the Ethereum’s values at all.
Your observation matches the data precisely. A median VQS of 1.8 means most delegates show rapid, uniform voting patterns — consistent with “paid advisor” behavior rather than independent governance. When deliberation scores are low and independence scores are low simultaneously, it suggests delegates aren’t critically evaluating proposals individually. The question is whether Lido’s governance structure incentivizes this behavior, or whether it’s a cultural pattern that could shift with more transparency around vote quality.
1 Like
How fun that it started with interesting data, ironically this:
and then we all realised the unmistakeable pattern of AI-generated slop…
Nevertheless, @BCV , you say:
You think there should be “tracks” on governance? As in some votes that have to do with infra, some with DeFi, and Delegates should only be voting on whatever their expertise is? Meaning, a Defi expert delegate should vote in the new Lido Earn proposals, but not on Lido v3 deployments? Honestly I think this is tricky and not all the proposals are very clear cut… most of them are actually pretty intersectional and have a, to quote @Jenya_K :
I think knowing about the financials, the structure of Lido Labs, having followed successful initiatives like CSM and less successful ones like SDVT (don’t shoot me!), having gone through a couple of GOOSE exercises actually gives a lot of value, I don’t know if more or less than being a “track” expert.
4 Likes
Fair point on the AI-generated text, @Lanski — the analysis narratives are AI-assisted, but the underlying data is real: 4,279 delegates, 24 DAOs, all sourced directly from Snapshot.
The discussion you’re having here actually gets at something our VQS model tries to measure with the “Focus” signal — do delegates vote within their expertise, or across everything? Your point about Lido governance being “intersectional” is exactly why a simple track-based model might not work. The question is whether we can measure vote quality without prescribing structure.
Curious what the actual delegates here think — does a Focus score even make sense for a DAO like Lido where context cuts across all categories?
I believe there is some confusion within Lido between two distinct concepts. In its current form, Lido delegates function more as an advisory subgroup or service provider under Labs, and they are rewarded for their contributions.
However, the key point I want to emphasize is this: as the primary staking intermediary on Ethereum, Lido should have a decentralized delegate set that acts as an ownership and oversight layer. This is crucial if the Ethereum ecosystem is expected to rely more heavily on Lido. At present, the protocol remains significantly centralized around Labs, without meaningful oversight or supervision, despite having a governance framework capable of supporting it and despite Labs’ stated commitment to transparency and accountability.
Establishing a decentralized delegate set would improve the protocol’s sustainability and transparency, and help position Lido as a neutral infrastructure layer on Ethereum. In turn, this would increase trust within the staking community. While dual governance strengthens the protocol’s security, I believe a decentralized oversight layer is equally vital in the long term.
What would the role of such a decentralized delegate set look like?
Lido operates under an optimistic governance framework in which Labs serves as the executor, while delegates primarily play a supervisory role. This structure requires far less technical complexity than protocols like Aave, where delegates effectively act as executors. As a result, Lido does not necessarily need a highly technical or centralized delegate group to function effectively.
How could a decentralized delegate set be formed?
Ultimately, the decision lies with Labs. If they genuinely want this layer to emerge, token holders must be incentivized to delegate their tokens, and delegates should receive commissions tied to those incentives in exchange for active participation. Otherwise, the delegate set will remain a centralized subgroup under Labs, and the protocol will continue to revolve around it.
Most importantly, any incentive mechanism should be credibly neutral and hardcoded, rather than subject to discretionary evaluations by a committee.
This is a really important distinction, @BCV — delegates as supervisory oversight vs. delegates as executors. It explains why Lido’s VQS looks so different from protocols like Aave.
Your point about incentive design being “credibly neutral and hardcoded” rather than committee-evaluated is something we could actually measure over time. If Lido moves toward a decentralized delegate set, tracking how participation, independence, and power distribution evolve would show whether the structural change is working or just cosmetic.
Right now our data shows Lido’s delegate set is small and centralized — which aligns exactly with what you’re describing. The question isn’t whether the current delegates are doing a bad job, it’s whether the structure itself creates enough oversight.
Appreciate you laying this out so clearly — this kind of context is exactly what makes governance data meaningful rather than just numbers on a screen.