Sushi Referral Program


It is proposed that Sushi implements a two-pronged referral program, aimed at retail users and institutional users respectively.

This proposal should be assumed to be a raw starting point and a terrible idea as default. I believe in our community to help shape this into its most effective form.


It is proposed that Sushi reward wallets - both individual and institutional - for onboarding new users onto the protocol.

To mitigate the potential for abuse, it is proposed that this be done via two separate programs:

  1. “Sushi Trusted Partners”: A manually whitelisted, traditional ‘% of added revenue’ model
  2. “MEOW-llionaire”: A gamified points-based leaderboard, open to all

Moreover, it is proposed that a Referral Reward Pool be created to fund both these efforts.


Referral programs are typically a no-brainer user acquisition tool in many traditional companies: users are rewarded with a slice of the revenue that they acquire for the company or, in this case, the value that they accrue to SUSHI/xSUSHI holders.

This approach has been largely underutilized in crypto, with some notable exceptions (i.e. Lido using theirs to increase TVL by ~50% and stakers by ~100%).

The aim is to increase the k-factor of the protocol - the number of additional users each user acquires - ideally to >1; this would scale growth, and value to SUSHI/xSUSHI holders, exponentially.

Specification A: Reward Pool

In order to avoid funding via fee revenue directly, it is proposed that a Referral Reward Pool be created to fund these two programs.

Moreover, it is proposed that this take a phase-by-phase approach, so as to expand the scale of commitment in line with proven efficacy. Ideally progression to the next phase would only occur upon community sign-off. An example of what this could look like:

  • Phase 1: X SUSHI (initial trial funding)
  • Phase 2: 10X SUSHI
  • Phase 3: 100X SUSHI
  • Phase 4: Uncapped until community decides otherwise

Reward Pool: Discussion Points
A1. Format (number and transition between phases, non-phase approach etc.)
A2. Funding allocation amount
A3. Alternative to Reward Pool

Specification B: Sushi Trusted Partners

Essentially a reskinning of Lido’s Whitelist Mode. Namely:

  • Only community whitelisted wallets are allowed to participate (focus on known dev teams, institutions)
  • Whitelisted wallets are rewarded with an amount from the Reward Pool equal to a portion of the fee revenue they acquire for the protocol through onboarded users

Sushi Trusted Partners: Discussion Points
B1. Pre-approved candidates
B2. Portion of revenue awarded

Specification C: MEOW-llionaire (working title ^_^)

This program would approach the other half of the coin: abuse-prone, non-institutional users.

Basic UX:

  1. Referrers create a unique referral link attached to their wallet address
  2. Referrers onboard Referees onto Sushi via this referral link
  3. Referrers accumulate points via Referee usage over a period of time known as a ‘Season’
  • Various NFT-based game mechanics are layered on at this stage
  1. Points place you dynamically on the Referral Leaderboard
  2. Rewards determined by placement at the end of a Season
  3. Rewards - in the form of MEOW tokens and NFTs - can then be claimed by Referrers
  4. Commission from NFT resale goes to refill the Reward Pool

Full proposed details (abuse prevention, detailed game mechanics, rewards) are laid out in this doc: MEOW-llionaire Details - Google Docs

MEOW-llionaire: Discussion Points
C1. Non-stupid name
C2. Other sybil-resistant approach to reward non-whitelisted wallets
C3. Variations on game type/mechanics

Many thanks if you read this far. Let’s get the discussion going! :point_down:


Referral programs are great! And your three Specifications definitely reflect successful programs from Lido, 1inch, Aave, dydx, and the comments on Maki’s tweet etc.

A) Capping the rewards pool
B) Whitelisting participants, like wallets and dapp contracts
C) MEOW or NFT rewards, incl leaderboard

If C) is everyone who doesn’t get whitelisted as part of group B), then I’d simply cap the rewards pool for both.

On B): An alternative to whitelisting as means to prevent cheating, could also be retroactive airdrops, like these already happening: Sushi Vesting for Yield Aggregators (SIMP #3)

On C): If you reward non-whitelisted users with MEOW dust and NFTs, I’d also include a reward for visiting the site daily - similar to Coingecko candy or Zerion xps. Simply because you got most of the UI and logic already, and you can optimize another important variable: The referrals to grow user acquisition and the gamification to increase retention.

On question C2: You could verify a Twitter account’s ownership of their ETH address, eg. through the Twitter API. That’s how Linkdrop does it, see this NFT airdrop.) You could also verify that they tweeted a certain text, or verify a certain follower number.
Or you could use a more privacy-friendly flow similar to Keybase, through Ceramic.


I think a referral program is a good idea.

Leaderboard is an excellent idea. Make it competitive and fun.

Would there be a way to reward referrers based on the amount of TVL they bring into the system? Or, can we only do set amounts?

If someone brings in a big “account” or a whale, reward for doing so should be relative to size (similar to commission model).

Whatever is decided on this proposal, I would agree with continuing this as an onboarding initiative to bring more value for the existing community and protocol and help those to understand what Sushi offers.


Absolutely support a contribution-reward tether.

The institutional program as proposed would reward $ directly corresponding to a % of fee revenue (so volume-based vs TVL); but, this formula could be modified ad hoc to whatever the community wished, but it would be less inherently ‘revenue negative’.

With the retail program as proposed, this is much easier. Point rewards can be modified frequently and assigned differently based on a strategic needs (for example bumping points for Shoyu usage during its launch).

An e.g. from the linked Google doc:

1 Like

I really like where you’re going here with this. Especially the the points-based leaderboard. It sounds fun and competitive for whoever would like to join.

I would propose including active-user onboarding as a factor to take into account in the points based system for non-institutional participants. I believe that there’s great benefit to adding active users even if they don’t bring a lot of revenue given the network effects that could have (those users referring other users). Obviously revenue/fee related compensation is more direct in rewarding the value added in direct cash flows, but maybe it’s missing this other component to compensating for adding value to the network through more users…

Either ways, I love this idea and think we should pursue referral programs and other incentives to spur platform growth through our community efforts.

1 Like

I like your idea on this…

Almost like a multi part rewards system. Could be rewards on the following:

  1. Largest single user referrer - so $ amount for one user who brings in a whale.
  2. Largest $ amount for a single referrer - for the person who brings in a lot of valuable accounts and gets to a big number.
  3. Largest number of users onboarded by one referrer - this speaks to the person who may bring in a bunch of small users but really has the ability to expand the network effect and build community.

I love the idea of a referral program, especially along the lines of your guidelines. Do you have an idea of what it would take for the community to build this out? Man power and Cost?

Referral programs are great for something like Lido. With Sushi, I’m not as sure… the great bulk of Sushi revenue is from the AMM (for now). And I see traders, especially large ones, increasingly shifting to aggregators (1inch, Matcha etc) in order to get the best prices.

So would investing in a referral program be as effective bang for the buck as, say, investing in extra yield farming rewards for Trident concentrated pools to drive more liquidity and thus more volume? I’m not sure.

Maybe Sushi can build out its own aggregator in which case this becomes a moot point. And maybe referral programs could be useful for products outside the AMM. But at least for now, the AMM is the biggest revenue driver, by far.


Nice insights, and great point! Both programs would have a different target groups, but looking at the bigger question of how to spend team time effectively, it would be nice to have a ROI on both, the ref program for the longtail of users vs. increased yield farming.

We could eg. analyze other DeFi projects ref campaigns and yield farming.
Did Sushi ever work with tagged onchain data, like Nansen’s?

@Unagi How do you plan to analyze data of say ref campaigns of various DeFi projects? Is this data open? You’re right, the target groups are different, but in the end, if the goal is to acquire more users, feel like there might be better game plans we can design to improve COA and ROI. Read this on Alex Arca’s tweet: “The following week after the Bit DAO sale, the weekly unique traders grew 27%.” This was likely because of the sale, but if a single step like that can improve our acquisition, feel like we should study all this, a bit more before jumping in to create a larger implementation.

1 Like

Didn’t dive into it that deep, and I guess it wouldn’t be as easy to track as impact of yield farming, since not everything happens onchain. But definitely worth a look.

Good point well made.

Fortunately in either ‘prong’ rewards can be set dynamically, i.e. we can lower rewards up to the point that the ROI makes sense. Granted, this is tethered to efficacy: lower rewards will result in lower usage.

In the extreme, this would reduce down to just the retail version with only NFT/gameified rewards, with zero treasury erc20s on the line.