Is Snapshot off-chain voting working for DAOs?

On-chain voting is in the DAO spirit but expensive. So many DAOs prefer to vote off-chain, most commonly via a service called Snapshot, which is set up to offer a quick and comprehensive experience that is very close to DAO voting. But is it?

Some stats

Snapshot has over 13k “spaces” using it, aka DAO communities. 10k of them on Ethereum, 1.5k on Polygon, 1.3k on the BNB Chain, and so on. And they’re certainly active: nearly 29k users authored over 84k proposals, and over 1.7M voters cast nearly 10.4M votes.

The Optimism Collective and the STG DAO are the most active spaces, each with over 1M votes cast. Meanwhile, in terms of the sheer number of proposals, PancakeSwap leads the pack with over 1.7k proposals.

As far as voting trends, there is certainly overall growth in voting but it’s not entirely linear, with both pullbacks and monthly voting spikes happening, prompted by the activity of specific DAOs (launches, major proposals) and by the crypto market in general.

How Snapshot works

Snapshot uses a combination of Web2 and Web3 technologies to make the process as transparent and cost-effective as possible. To that end, it has several components:

1. The user-facing Snapshot UI, which is free to use but does require people to sign in with their Web3 wallet to vote or propose.
2. The Sequencer that makes sure all interactions are recorded in order and any manipulation can be traced (Snapshot’s version of blockchain transparency).
3. Voting and Validation strategies to give projects and their users many options to vote (via ERC-20 tokens or ERC-20 NFTs, for example) and validate eligibility (like checking if the user holds a particular POAP).
4. Score API to calculate how much voting power each user has
5. And the Data layer to store the voting history on their centralized database with a reference address on IPFS.

Case study: PancakeSwap’s identity crisis

PancakeSwap presents an interesting Snapshot case study not only because it’s the most active DAO there by the number of proposals, but for another, very surprising reason. DAO’s create spaces by providing their reference ENS address, which PancakeSwap did (cakevote.eth). However, there is another space on Snapshot with the name Pancake and the exact same logo. It even has more proposals than the real PancakeSwap. Savvy users can tell the real one apart based on the white checkmark in its name. Yet, how many users have been fooled into using the fake (or at least, unofficial) one?

A hint at the second one’s agenda can be seen in its proposal descriptions:

Question is, how does this serve the PancakeSwap DAO? A quick look at the real PancakeSwap space shows that they’re not too happy with Snapshot. Just read the currently active “Proposal to enhance governance proposal system” where 1.7M users expressed their grievance via voting (admittedly, this vote only had one option). Indeed, much of PancakeSwap’s massive number of proposals is due to those proposals being… entirely irrelevant to PancakeSwap.

Case Study: Optimism Collective

Contrast that with the Optimism Collective’s Snapshot space, which has by far the most votes cast at over 1 million, yet manages to keep proposals on-topic. Its number of proposals is also a more reasonable 93 rather than over a thousand.

The best sign of Optimism space’s voting health is that the voting is not near-unanimous, showing a healthy divergence of opinions and a level of proposals with enough at stake to cause such disagreement.

On the other hand, it is still hard to understand which of the 93 proposals are of higher priority. Even filtering by “core” shows the same proposals.

Snapshot: worth it?

Snapshot’s popularity is evidence that it is a good service and a popular one among current DAOs. Among its advantages are the lack of a gas cost, the simple UI, the many customization options for voting and validation, and the ability to replay every action in a public way in order to detect any malicious attempts.

On the negative side, it’s not actually decentralized nor native to the DAOs. It’s hard to find the truly important proposals and — as PancakeSwaps demonstrated — easy to get confused by impostors.

One important aspect of Snapshot is that it does not actually execute the voting decisions, leaving that up to the DAOs themselves. There are integrations for on-chain execution of Snapshot proposals to go around that. Still, at the end of the day, it’s up to the DAO members with access to the multisig to make the decisions. And that is dangerously centralized for DAOs to rely on long-term and for important governance decisions. Certainly not in the age when exploits happen weekly and there are people’s money and trust at stake.

Now what?

Off-chain voting’s time and cost savings are important. As such, Snapshot plays a valuable role in the current DAO mechanics.

Still, DAOs must have reliable access to an in-house tool for both on- and off-chain voting. PancakeSwap’s users are frustrated with all the irrelevant proposals. That’s why a clear system for bringing worthy votes up to the top is important. Why having different settings for different vote types helps users to focus on what’s important without getting frustrated with the system. Those same frustrations experienced by DAO members that have to use Snapshot are what we were concerned with when designing DeXe’s DAO constructor. DAO tools of the future will help DAO members not only vote but know when and why to do so.

Stay tuned!

Website | Telegram channel | Telegram chat | Facebook | Medium| LinkedIn | Twitter | Reddit| Discord