Introduction of Transaction Fees on Front-end Market Orders to Support Protocol Growth

[bozhangles]

Summary

This proposal introduces a transaction fee on Market Orders executed through the Odos front-end to generate consistent revenue for the DAO Treasury. The goal is to support the long-term sustainability and growth of the Odos protocol, reduce reliance on unpredictable positive slippage revenue, and align incentives through the Odos Loyalty Program.

Motivation

Currently, the Odos protocol does not charge a standard transaction fee on Market (Advanced) Orders, instead relying primarily on positive slippage as a revenue source. However, this model:

• Is statistically unpredictable and difficult to scale
• Does not provide stable funding to grow DAO initiatives or improve protocol infrastructure

To address these limitations, the DAO proposes to introduce transparent, minimal fees that can serve as a reliable revenue stream for the DAO Treasury while ensuring continued best execution for users.

Frontend Fee Structure

• Volatile Asset Market Orders:

  A 0.15% (15 bps) transaction fee will be applied on swaps between volatile assets

• Stable Asset Market Orders:

  A 0.03% (3 bps) transaction fee will be applied on swaps between stablecoins or stable pairs

Note: Limit Orders and Protected (Simple) Swaps will not be affected by this proposal

Loyalty Program Rebate Eligibility

• All fees collected from Market Orders will be eligible for rebates through the Odos Loyalty Program (with the docs being updated to reflect this)
• Rebate tiers will be based on user activity and token ownership per Loyalty tiers
• This mechanism ensures that power users and loyal contributors are rewarded and incentivized to keep trading through Odos

Revenue Allocation

• Collected fees will be directed to the Odos DAO Treasury.
• Funds may be used for:
  • Protocol development
  • Liquidity incentives
  • Partner integrations
  • Marketing
  • Buyback and other rewards (via future DAO proposals)

Benefits

• Establishes a predictable, scalable revenue stream for the DAO
• Reduces dependence on positive slippage, which varies with market conditions
• Enables the DAO to fund meaningful initiatives without excessive token emissions
• Creates a clear alignment loop between users and the protocol

Next Steps

• Upon passing this proposal, the development team will implement the fee mechanism in the front-end routing logic.
• The Odos Loyalty Program and related documents will be updated to reflect fee rebate eligibility and rebate tiers.

[pomme]

This is effectively a fee on loyalty. Loyal contributors will not be rewarded; instead, they will be fined 0.03-0.15% of their trading volume and, at best, compensated with only half of that on the highest Loyalty tier. This does not incentivize them to keep trading through Odos.

The proposal is either incompetent or intentionally misleading.

Users are loyal only because of competitive prices, and this proposal undermines that advantage. The most active users with the largest trading volumes will leave Odos, because these fees hurt their profits, leading to a significant decline in revenue.

This proposal does not support growth, but rather the long-term decline of the protocol. Soon, there will be nothing left to scale or improve. The protocol will be dead, and reverting this proposal later will not restore trust—once-loyal users will leave and forget about the protocol forever.

I suggest we think twice before making an irreversible decision.

It’s clear that most users do not read these proposals or vote, so a slight administrative push can get any proposal approved. I am not delusional about that. But I am addressing those who built the protocol, those who are not planning to rug pull, those who care about its future. Don’t fool yourself—Odos offers nothing unique except routing. There is nothing special about simple swaps, limit orders, gas fee rebates, or anything else. Price is the only competitive advantage. Be sure that users can read the prices, and they will leave if Odos is no longer the best.

I am not suggesting we negotiate to lower these loyalty fees; I suggest we reject the proposal entirely. Also I suggest the Treasury administrators make a reasonable effort to demonstrate their commitment to common sense.

[Slothman]

I’ve been a long-time ODOS user, and as someone who uses it almost exclusively for swaps, this new fee structure will push me to switch to another aggregator immediately. The proposed fee is simply too high. I suggest reducing it significantly to around 0.02% for volatile pairs and 0.01% for stable pairs, or something in that range

There’s no real “moat” here that justifies such high fees, especially when other aggregators offer similarly effective routing with no fees at all
Not all users can be expected to buy into the ODOS token just to make the fee more palatable that model isn’t practical

[bozhangles]

You raise a great point: traders aren’t loyal. I myself am a trader, and I primarily uses bots to trade, so I get this firsthand. In fact, bots account for nearly 80% of all Ethereum transactions today. Because of this dynamic, your point cuts both ways.

We’ve conducted financial modeling that shows even with a 70% drop in volume, revenue from modest fixed fees would still exceed what we’re earning today. This gives us the financial room to expand headcount and invest in product development, platform expansion, and tooling - enabling long-term growth in other directions.

It’s also worth clarifying that Odos offers more than just “routing.” Our routing engine is industry-leading, consistently delivering better pricing than most competitors. That’s why Odos is the top SOR on Base, Sonic, Avalanche, and more.

But I agree with your broader point: this is a cost compression game. Over time, everyone races to zero fees to win flow. And once revenue hits zero, you can’t pay engineers, market the product, or iterate—at which point you’re essentially dead. We’ve seen this fee dynamic play out in other industries, like ETFs. With ETFs however, they’re now making their revenue from securities lending. We can’t do that.

So our strategy is simple:

• Implement competitive fees (still lower than many peers),
• Increase revenue significantly, and
• Reinvest that revenue into building sustainable differentiation beyond the routing we’re doing now.

And if volume ever drops off a cliff? Given trader and bot behavior, we expect we can reverse the fees and recover most of the volume, as switching costs are minimal and we’d still be the best routing platform in many cases.

For context, even with those fees, we’d still be quite competitive in the industry. Here’s how other players approach fees, and they’ve done very well even with these fees:

• Cowswap: 50% of surplus, up to 1% of trading volume + 10bps flat on Gnosis Chain
• Uniswap: 30bps flat + LP fees
• Hashflow: 1–2bps flat
• Jupiter Aggregator: 5–10bps flat
• OpenOcean: 10bps flat + positive slippage sharing
  and the list goes on.

Our view is that fee compression is inevitable - but it’s also an opportunity that we can take advantage of. The protocols that act early, monetize while we can, and expand in other fee generative areas will win this cycle.

[pomme]

So, you’re suggesting we burden human frontend users with fees — users whose actions are constantly driven by irrational thoughts, emotions and sentiment, who are always reevaluating the trustworthiness of the product, and who are the sole target of any marketing campaign — while at the same time letting the bots, which make up 80% of the total volume, off easy? Am I understanding this correctly?

Is this because, according to your own words, you primarily use bots to trade, or is there some hidden logic that I don’t understand?

I’d like to think you understand how terrible a 70% drop would be, but you wrote about it as if it’s not a big deal. Any artificial drop in volume is devastating. Statistically speaking, a 70% drop in volume means roughly the same 70% drop in everything: number of users, feedback messages, mentions on social media, and bug reports. Does your financial model account for how valuable a committed user is? If this proposal is agreed upon, we’re going to lose a lot of them. Look, @Slothman — a committed user who even writes messages on the governance forum (who ever does that?) — is leaving immediately.

This DAO shouldn’t be all about money. Let’s be people and remember about the community, not just the revenues. The opportunity you want to take advantage of hurts the very DAO members you expect to vote on this proposal.

Also, @bozhangles, could you please clarify on whose behalf you’re speaking? Whose strategy and whose views are you representing?

[Nuriag]

Thanks @pomme and @Slothman for your input, this is exactly the kind of conversation DAOs are meant to foster. It’s important that we raise these questions and engage the community.

Just to add a bit of broader context:

This isn’t simply about adding a fee. Behind the scenes, the team has been working hard on meaningful improvements to the protocol, from routing algorithm upgrades to new features designed for more advanced traders. Some of this was already shared in the last AMA with Aaron, Head of Product at the DevCo, and we’ll be sharing more updates soon.

The goal is to ensure that even with a small fee, trading on Odos continues to deliver best-in-class performance with faster execution, better routing, and stronger overall outcomes compared to other aggregators. So this isn’t a fee for the sake of it; it’s part of a larger strategy to make the product stronger.

We’re also preparing clear comms to explain what’s changing and why - with examples and comparisons to help users understand the impact and the value being delivered.

It’s true that introducing fees is a sensitive shift. But relying solely on positive slippage as a revenue stream limits what the DAO can plan and build. That’s why the proposal also includes a Loyalty rebate system, allowing active users to recover part of those fees based on their activity and engagement.

And just to emphasize: @bozhangles is ODOS DAO Treasury Lead and has worked very hard and closely with the DevCo team analyzing real data, building models, and exploring different long-term scenarios. It’s been a serious, thoughtful process rooted in sustainability.

Thanks again for raising these points.

[eliteness]

On the first look, we would be very much against something that causes quote: 70% drop in volume.

But coming from the Real Yield background, we are among the ones who openly advocate for generating actual real revenue for the protocol. These fee are a necessary evil that we need to have.

With that said, 0.15% is definitely a wrong number.

We suggest starting with something like 0.01% on initially, and then re-evaluate its impact on userbase every 3 months, and gradually increase if it has a positive or neutral effect.

Considering that the change in the front-end is simply about Replacing the Referral ID while querying the SOR API, it is extremely easy to update the fees frequently.
Another change required in the Front-end would be to responsibly show the new Front-end Fees for transparency.

As such, we recommend the proposal be thus modified to a Tangible, Deterministic, Continuous evaluation of (say, monthly) Active Users on Front-end in a Cycle and the Front-end Fees raised during a Cycle.
It is extremely important to balance Fee-adjustments with the On-chain activity. In low activity periods, an extra fee is viewed by users as a Penalty, as rightly pointed out by pomme. While during high-activity periods, it would be much more forgiving to have some fees. In either cases, a small baseline front-end fees is pragmatic to ensure Odos lives to see the next Decade of DeFi.

[pomme]

If we are not rejecting the proposal immediately, I can suggest some constructive ideas as well.

I totally agree that the suggested fee rate is dangerous for the very existence of the protocol. @Nuriag, this is exactly the kind of warning this forum is meant to foster. The treasury is recklessly pushing a very risky move. Every Odos user and every DAO member is at risk. I cannot be more concerned. When things go wrong, it will look very bad for everyone involved, because all the warnings were there. The fee rate is so extremely high that I even suspect it was chosen intentionally to demonstrate that the DAO treasury can make compromises when the suggested in the proposal fee is eventually lowered.

Doubling the fee rebate at every tier of the Loyalty program (up to 100% at the highest level) might be a real compromise, which would have a huge impact on the appreciation of the $ODOS token.

It should also be mentioned in the proposal what the base for the fee will be; this should not be left to the technical team to decide.
If negative slippage occurs, the base for the fee should be the actual output amount. If positive slippage occurs, the base should be the quoted amount.
Additionally, I suggest that in the case of positive slippage, the fee should be reduced by the amount of positive slippage, down to no fee if the positive slippage exceeds the fee. It is simply not fair to take both the positive slippage and the fee.

[Pull]

I agree with the restraint in initial scaling and most of what you wrote. For me the most important piece of this proposal is the ability to track its success/failure in real time. You mention:

As such, we recommend the proposal be thus modified to a Tangible, Deterministic, Continuous evaluation of (say, monthly) Active Users on Front-end in a Cycle and the Front-end Fees raised during a Cycle.

I want to stress that active users is the wrong metric to track here and we should be looking at revenue generated from the router contracts. We can easily normalize this for chain specific gas price volatility (which is extremely correlated with price volatility / DEX volumes). These are both queryable and act as better metrics we should be optimizing for.

My recommendation is to create a source of truth for monthly revenues (swap, limit order, FE fee, etc) that can track historical and current data. Then normalize this based on the chain specific gas pricing. If there’s measurable decline in “similar gas volatility conditions” then we should look to remove the fee

[bozhangles]

Thank you - I’d like to emphasize that the 70% drop is the worst case scenario. It’s more likely that we will remain very competitive because competitors’ fees are typically higher. In addition, we have several updates coming soon (cross chain transactions, V2, etc.) that will allow us to retain users and deliver value. So as a result, this number isn’t the one I could anchor one, but was more shared to give an illustrative example that even amongst this type of scenario, the fees would be significantly value accretive to the protocol.

@pomme the drop in marketing metrics isn’t the right way to view it because as I mentioned, the majority of our traffic comes from API users. Whilst we do want the DAO to focus on the users and community, at the end of the day, it’s important to remain pragmatic about the direction of the space and that to ensure the long term survival of the DAO, having a set of fixed fees makes sense (as echoed by many of the other folks here).

I’m surprised by the thinking that the fee is so high @eliteness @Pull given what many competitors charge (0.1% up to 1%). Why do you think it wouldn’t make sense to charge more, but still competitively priced relative to competitors?

[Pull]

In my experience, users typically switch to a new FE for one of two reasons:

1. The expectation of an airdrop (potential yield)
2. A meaningful difference in fill quantity (guaranteed yield)

The sensitivity of #2 is harder to gauge, especially since many FE users are uninformed flow and may not even notice or care about the presence of a fee.

Personally, I don’t see this new fee as a major concern. Still, the only way to understand its real impact is through careful, incremental rollout. Opportunities to test changes at this scale are rare. Most teams pay advisory teams just to be told, “If you raise the FE fee, you’ll earn more” often with a six/seven figure invoice attached

[bozhangles]

Yes, agreed. We don’t expect this new fee to be a major concern either. As we think about this though, I think framing is an important part of the discussion. For example, if we had initially said we should implement a 30bps fee, I think everyone here would’ve been happy with moving down to a 15bps fee.

The main issue with scaling up from my perspective is the perception of oh wait, you’re going to increase fees on me? And the emotional reaction towards that. Setting set fees for the first time is understandable I think to most people - as you saw Tron is IPO-ing as well. I wouldn’t be surprised if the Uniswap leadership team is thinking about whether or not they should do that as well. As more crypto companies go down this route, there will be a greater emphasis on revenues and margins. Companies with strong revenue bases and margins will make it, companies that don’t, won’t. So implementing fees broadly makes sense.

Scaling it will be tricky - you saw the visceral reaction that people had when Netflix raised fees for the first time. They were perfectly happy to pay the whatever it was, £9.99? But £10.99? People were fuming. So I think framing in this context will be important as well. Would it be better to set fees once, and say that’s it? Or do we set fees, and then increase the fees periodically. As I think about it more, I think the latter would actually result in more negative sentiment.

[pomme]

Don’t you think that Netflix conducted its research and that this modest increase is probably optimal? Did ODOS DAO conduct any research on the impact of the proposal, or did it just look at the fee rates of competitors whose situations are not the same as ODOS’s?

I suggest suspending the vote on this proposal until such research is done.

Also, let me remind you that ODOS DAO is a non-profit organization, while Netflix is not. Most Netflix users are not shareholders, whereas almost every ODOS user is a DAO member.

[bozhangles]

@Pull I don’t see the question you dropped onto Discord here, but we’re trying to keep the feedback to the forums so lets bring the convo over here.

What you’re seeing is positive slippage (PS), a byproduct of volatile markets where the executed swap amount is slightly better than quoted. Historically, this has been collected by the DevCo, which is the protocol’s core service provider. These funds cover all operational, infrastructure, and development expenses (e.g., headcount), as the DAO currently does not fund these directly.

This PS is inconsistent, market- and volatility-dependent, and cannot be relied on as a stable or long-term revenue stream - hence why we have this proposal here. Moving forward, with the expansion of the DAO, more of the PS (as well as some of the other fees, like this mentioned in this proposal) will gradually transition to the DAO treasury by the end of 2026.

To be clear, all revenue generated by the protocol today is netted against service provider expenses, as the DevCo (the service provider) is paying for all the headcount, infra expenses, etc.

[Pull]

For the sake of record keeping heres the question

" I created a model for estimating ODOS revenue and noticed a significant discrepancy compared to the new dashboard. For example, the router frequently calls Swap Router Funds to transfer assets to a safe. For instance on Base, that safe is: 0xa7471690db0c93a7f827d1894c78df7379be11c0 . I’d appreciate clarity on these value flows and their end use. Notably, actual revenue is much higher than what reaches the DAO safe. Specifically what are these funds currently being used for and what future plans exist for these funds?"

[eliteness]

Hello!
Eliteness.Network contributed the original code for the DefiLlama adapter of Revenue & Fees generated by Odos Protocol. You can track those metrics here:
https://defillama.com/protocol/odos?groupBy=monthly&revenue=true&dexAggregators=false

You can also view the Treasury addresss we marked and can visualize the inflows and outflows using any advanced trackers online.

According to it, Odos roughly earns $350K monthly. We request fellow DAO members to cross check and verify these numbers. It would be lovely to have a breakdown of how these funds move, and what actions we could expect from these collection addresses in the future.

As for the “Fee Number too high” issue, it is vital to understand a few key patterns:
Consider the fact that any uniswap-v2 like old generation protocols (sushi, pancake, curve) have the default pool fee of 0.2% to 0.3%.
With an extra 0.15% Fees, users would suddenly experience an increase of 75%.
Many of the trades happen on Blue-chip assets, and usually the fee-tiers are quite low. ETH/USD, the most traded “volatile” pool in DeFi, has fees-tiers of usually 0.05% or lower. Making users pay 4x suddenly could be a trust-shock.

Again, we would advocate for a Continuous & Comprehensive Evaluation for this specific parameter. With a Long-Term product timeline in mind, it is best to introduce such taxing measure mildly instead of abruptly.

Our Recommendation:
Start with 0.01% introductory Front-end Fees. Thereafter each month have a comprehensive evaluation of the market trend and effects of fees on Odos Frontend usage each Month.
Thereafter, have a vote to either INCREASE by 0.01%, DECREASE by 0.01% or leave it UNCHANGED.

[pomme]

I have some ideas for a compromise.

I’ve noticed that some so-called agnostic to Odos users are planning to turn to external wrappers like DefiLlama Swaps to avoid the new fee. However, this opens up the possibility that they may be lured into using other aggregators available there, which would decrease the amount of positive slippage Odos collects.

I suggest that Odos fork its frontend to create an alternative version without this new fee. It is common practice in DeFi to have different frontends for the same protocol. This fork would not be updated or maintained beyond its basic functionality. The link to it would be buried deep in the documentation and not advertised in any way. We could even prohibit advertising it on the Odos Discord. This way, new users will use the fee-enabled frontend as intended, while long-time users who dislike the new fee and got into the habit of using Odos without fees won’t be abruptly cut off from the Odos ecosystem. They will also have a chance to explore the benefits of the new updates and gradually adjust to the new fee structure. Personally, I think this would be a very appealing gesture that shows Odos cares about every single user. Although I believe its maintenance will be extremely low-cost, we will discontinue support for it if the additional positive slippage it generates drops below the cost of maintaining it.

Another idea is to separate the fee into two proposals.
The first would be a more modest sustainability fee, which would add some predictability to revenue and fund only protocol development and maintenance.
The second would be a more ambitious growth and expansion fee to fund everything else.

Many vocal users who oppose the proposal agree that a small fee is necessary. However, with a single proposal, it’s all or nothing: either the full fee or no fee at all. If the voting isn’t manipulated, the outcome is always unpredictable.

This approach offers a good compromise for both sides. Supporters of the fee can secure a modest fee, while opponents have a better chance of rejecting the ambitious fee.

[bozhangles]

Hi there! Unfortunately our legal council has explicitly told the team that full details between the Devco and the DAO should not be shared (same reason for example why a fee contract between JPM Asset Management and JPM IB should not be shared / made public), but I can try to provide as much color as I can without hopefully ruffling any feathers.

Previously, all of the positive slippage that was earned went to the Devco, as the Devco covers head count, operational expenses, infrastructure expenses, etc. Currently, with the DAO team in place, a small portion of the positive slippage is now going into the DAO for the running of marketing initiatives and growth campaigns. Moving forward, the goal is to make it more evenly split and eventually, have all headcount, operational costs, etc be migrated under the DAO. This would allow for the maximum level of transparency, and is something that we’re taking the steps to do. Just needs to be carefully planned out taking into account of the shifting regulatory regime dynamics.

Moving forward, we would expect a portion of the assets in the collection addresses to move to the DAO treasury, and a portion to continue going to the Devco. The split between here will shift more in the favor of the DAO over time.

With percentages, I think we have to be careful of framing here. Because the fees on DEXs were already low to start with, so any number would seem quite big percentage-wise. My question would be, since many of our competitors already charge much higher fees, wouldn’t it behoove us to increase fees to a point where we can generate consistent and meaningful revenue as opposed to 0.01%, which won’t move the needle for the company revenue-wise but will still result some backlash?

Using Cowswap as an example, they have fees up to 1% of transaction volume. That would be 566% higher than what we would be charging, and they still process about $200mm of ADTV, and are at $84B of lifetime transaction volume. Their annualized revenue is on track to be $17.8mm this year vs Odos of $7.5mm. So I think we’re being overly careful about how sensitive traders are to fees, especially when looking at what is already happening in the market. Imagine how quickly Odos could grow with an extra few million in annual revenue - back of the envelop math shows that this would be very achievable. Attaching here a vastly simplified model using publicly available numbers to illustrate this point. Please note - this assumes ALL of the revenue is derived from volatile swaps, etc, when in reality the percentage derived from this is less. This should only be viewed as an illustrative example.

Using this very simple assumption, you can see that with 0.15% fees, we would experience an increase in revenue which would benefit all DAO members, the protocol, and the community. And as we’ve seen from Cowswap’s fees which are much higher than our proposed fees, traders can be quite fee agnostic up to a certain point.

Consistently raising fees until we find the “breaking point” of users I think would actually turn more users off than just settings fees once, and leaving it at that. After all, we are providing a valuable service - it makes sense to charge for it.

[bozhangles]

Thanks for the suggestion, but unfortunately maintaining two separate front ends, pushing changes to both front ends, etc, isn’t something that would make a ton of sense. Different front ends for crypto projects exist (using Synthetix or Liquity as examples here) because they’re typically managed by different teams and projects. Otherwise, having two separate front ends, one buried in the docs, isn’t quite industry standard.

Perhaps I’m missing something here, but I don’t fully understand the point of a sustainability fee vs a growth fee? If the protocol isn’t growing and expanding, then the protocol is dying. Suggesting that the protocol not expand and grow and just sustain its current state would be very detrimental to the community and the company.

[pomme]

You’re missing the point that this growth fee doesn’t assume any growth at all. In fact, there’s an opinion that it may actually have the opposite effect. Since we’re having trouble understanding each other, I suggest we put this to a separate vote, so we don’t risk rejecting something we all agree on.