Revenue Model for Instadapp

I think we can refer to the centralized exchange, turn Inst tokens into the common currency of instadapp, and set some lending projects to hold Inst tokens to participate, so instadapp will have a huge cash flow

A simple, good and fair income model!

Hey everyone, fantastic discussion so far. I personally think that the multiple revenue approach discussed above is great. I had some thoughts I’ll put below. I apologize for the length but I have been thinking about this and doing some writing over time. I’m going to get into some general revenue stuff but then a lot of tokenomics related to how the revenues are handled as I think having that hammered out before the revenue comes in is important.

First, I think the discussion of timing brought up by many in this thread is important. A possible middle ground is that new integrations can be given a large grace period before the “switch is flipped” for their modules and fees begin being charged. They will see an uptick in usage of their protocol because of the Instadapp integration, which benefits them, and later on the revenue will kick in for the INST holders. I think that something like a year gives ample time and incentive for both parties (the protocol being integrated & later the INST holders/treasury) to see the benefit. I think people bringing up how fees could damage growth is a very real possibility and careful assesment of what competetors are charging in fees is important.

The next logical step is what is done with the revenues and gets into tokenomics. I personally like the current tokenomics model laid out by Sushi and I think they are one of the only projects that has multiple revenue streams all benefitting the token holders / team.

When users make trades on the SushiSwap Exchange a 0.3% fee is charged.
0.05% of this fee is added to the SushiBar pool in the form of LP tokens for the relative pool.
When the rewards contract is called (minimum once per day) all the LP tokens are sold for Sushi (on SushiSwap Exchange).
The newly purchased Sushi is then divided up proportionally between the xSushi holders in the pool, meaning their xSushi is now worth more Sushi.

I think a way to meaningfully apply that model to one of the new products that is being teased would be in Samyak’s list

  1. Instadapp’s own flashloan (forum post coming soon): Inhouse flashloan of Instadapp protocol. Most deepest & biggest flashloan.

Based off this and what was said in the Gelato interview, if there was a native Instadapp lending market similar to Aave’s that launched the example transaction flow would look like;

  1. User 1 deposits USDC to Instadapps native lending pool.
  2. User 1 gets iUSDC. (Instadapp interest bearing USDC)
  3. User 2 borrows USDC from the lending pool and a small USDC fee accrues to the lending pool.
  4. User 1 decides to withdraw redeeming their iUSDC back for USDC and gets their piece of the fees.
  5. Instadapp then takes the accrued protocol fee at redemption and sends it to the treasury having User 1 pay the gas for both transfers as part of their position close.

What’s done with the USDC after that, in the example, is in all of our interest here and I by no means think this is the only right way, but I feel like the Sushi model is decently fair to all parties. It also would allow for a single staking mechanic of INST that would allow long term holders to earn more INST without having to participate in other ecosystems or open themselves up to IL risk via the LPs. Executing that in steps continuing on the above example I made would look like;

  1. All USDC in the treasury is batched and sent through aggregator/router like Matcha and used to purchase INST on some regular basis
  2. The newly purchased INST is deposited into the xINST (staked INST) pool

This creates a virtuous cycle because for people not staked, INST is being purchased benefitting them via positive price action, for stakers the pool of INST they can redeem their xINST for is increasing and directly benefits from volumes. There are other benefits from models like this too such as increased volume and it’s easy to integrate into future products because there is a single mechanism that benefits the team and holders rather than having to set different conditions on different products. It also would allow for the team to stake their INST or some portion of it and have consistent revenue tied to their labor rather than them having to time buys and sells to manage their expenses.

I hope that made sense and I hope to hear more ideas about tokenomics/revenue! I’m really excited to see what the team has been working on/teasing.

Also, on a side note I was also curious if anyone would be interested in forming an Instadapp Community Research Group where we can take different topics from the forum and go in depth. I’ll make a new thread down the line to gauge interest but my immediate thoughts were putting papers together comparing flashloan fees from other protocols and another paper looking at tokenomics models for other high TVL projects (maybe looking at UNI, SUSHI/xSUSHI, & AAVE) and highlighting the ups and downs of each model. Or the group could even go down Samyaks list above and do papers just looking at how revenue is collected by other projects for similar functions so that governance participants can have some relative idea what other projects are doing and what they charge the end user. Maybe the team already does this internally, in which case let me know. Cheers!

7 Likes

Hi Duo, this a terrific post, lots of great information and the ideas presented make a lot of sense. Just to add that the interview with Gelato you might have been referring to is this one entitled The DeFi Scoop Podcast with Dave Liebowitz. Cheers, Neal.

3 Likes

Appreciate your in-depth overview. Just looked into the Sushiswap’s distribution model, really exciting. Liked their idea to distribute it to the community.

They swap all their fees in Sushi and then distribute it to xsushi holders. Should we do apply the same model?
This modal will involve buying of INST and then distributing it to users again by any kind of rewards. One major difference between Sushi and Instadapp is Instadapp still has a huge treasury of INST to be distributed in the community over next 4 years while most SUSHI are already in the market. So does buying back of INST makes sense? Also, we realized having high rewards make people keep the tokens in the pool rather then use them in governance or else where so incentivizing users on voting could be a better incentivization modal?

Nonetheless, after seeing so much discussion around revenue. I think having a fee on swaps might be the best and easiest way to go considering most swaps (>95%) on Instadapp goes through strategies which also uses flashloan (which is free of cost) and complex composability. We can maybe have a vote and see what community thinks on this and if yes then how much should it be for stable coin and non stable coin pairs.

7 Likes

Really good points here and agree that some way to incentivize people to vote will go a long way. Agree that the pools currently keep a lot of INST away from voting. However, this program is scheduled to end and we can reevaluate then. That’s why I’m liking the swap fee as a first step if we want to go down the revenue path earlier than expected.

4 Likes

Hey Samyak, lots of good thoughts.

This model will involve buying of INST and then distributing it to users again by any kind of rewards. One major difference between Sushi and Instadapp is Instadapp still has a huge treasury of INST to be distributed in the community over next 4 years while most SUSHI are already in the market.

In Sushi’s case I believe that the xSushi pool does not receive any of the Sushiswap DAO’s SUSHI, it’s purely a pool that is used to capture fees and distribute them to stakers and allows the stakers to vote. So with INST, if there was revenue going to the DAO through fees, the revenue is pooled in the INST/xINST pool via the buyback mechanic. This allows the DAO/Team to use all of their INST to incentivize user growth (like the farming right now), pay developers and fast track integrations.

I have the opinion that separating the two funds the will allow for maximized growth because using the revenue to do the buybacks will also increase the value of the INST held by the DAO/Team. It will also create confidence in the platform/token beyond the 4 years because there will be a direct link for tokenholders to benefit for the entire life of the project once the incentives from the DAO slow down.

Also, we realized having high rewards make people keep the tokens in the pool rather then use them in governance or else where so incentivizing users on voting could be a better incentivization modal?

Yeah this is difficult for every project (generally there is low voter turnout in defi), I think it would be easy enough to include xINST in the snapshot for voting because it’s easier to track who deposited but when you have more complicated products like LPs it gets messy. Having the xINST be able to stake and vote may help increase voter participation in the long term because users will have somewhere they can stake their INST without risking it and be qualified to vote. (To mint xINST the user is depositing their INST so it should be fairly easy to track for governance.)

Nonetheless, after seeing so much discussion around revenue. I think having a fee on swaps might be the best and easiest way to go considering most swaps (>95%) on Instadapp goes through strategies which also uses flashloan (which is free of cost) and complex composability. We can maybe have a vote and see what community thinks on this and if yes then how much should it be for stable coin and non stable coin pairs.

Yeah, I was looking and it looks like MKR charges .05% fee for flashloan borrowing and AAVE charges .09% for flashloan borrowing. Are you picturing starting your own lending market or adding a small fee onto to the modules you built that access the other lending pools / swaps? If it is adding a fee to the modules does the team feel it would slower user growth/volumes or would it not matter ultimately?

4 Likes

Yeah! Makes sense. In long term using fee for buybacks and maybe distribute it again to community & INST holders could be a good move. Having a separate staking pool like xINST which will collect fees makes sense & interesting but still have to factor staking should not suck up all the INST liquidity from market considering at this early time. I still kind of like the idea of distributing fee to maybe INST LP pool so it also helps in better market liquidity of INST and will also make it easier for INST to get listed in DeFi protocols (Aave, Compound, Maker, etc).

Our flashloan will be different. Basically, you can flash borrow from any DSAs any position (Compound, Aave, Maker, etc). We can maybe start it with normal balances and move forward from there.

Not sure about slow down the growth if the fee is low it shouldn’t matter. Although, it could still be better to add more utility before charging any fee like Automated DeFi Limit Order, Automation to save from Liquidation, Guardian, etc which will provide more utility to users which will act positively on charging fee.

4 Likes

Some great discussion here!

Sushi takes a 5 bps fee from their liquidity providers, but they make up for this by distributing SUSHI to the liquidity providers. In this sense, one could argue that they are not that different from Uniswap. Sushi is essentially selling their governance token to liquidity providers over time, just as Compound is selling COMP to their borrowers who pay higher interest rates to get COMP. This leads LP providers who are choosing between Uniswap and Sushi to weigh swap fee volatility against SUSHI volatility.

Deciding whether to charge the users or the protocols seems like an interesting instance of the pricing problem in a two-sided market. One of the intuitions from the literature is that you charge more to the side who uses multiple platforms and subsidize the side that uses only one. For example, merchants accept multiple credit cards while many consumers use only one, so credit card companies actually pay the users to use their cards. This may imply that Instadapp should charge protocols more than users.

On the other hand, I agree with @Hillbilly_Chess that charging protocols may be labor intensive. One compromise could be to charge the users by default, but allow protocols to opt in to a user discount program. Users do not have to pay (or pay less) fees when interacting with protocols that participate in the program, and in return the protocol makes a recurring payment to Instadapp.

5 Likes

Some really good points raised here, it’s important that the logic is being checked.

The question I want to raise is whether or not it would be possible to charge fees anyway.

Any fee charged on even a small % basis would notably eat into users’ profits meaning that although Instadapp is convenient to use, it becomes arguably too expensive.

At the moment, the product is being given away for nothing and so there are lots of users passing their money through Instadapp, however as soon as using Instadapp reduces user’s profits many of those users will go elsewhere - ultimately all Instadapp functionality is available elsewhere, but not all in one place. Instadapp’s market proposition is one of convenience.

Therefore a fixed attractive fee for various operations could be used, with the emphasis on reducing those fees further for smaller users, enabling Instadapp to support wealth management and distribution for all - DeFi is not just for the wealthy, the crypto community deeply value their opportunity for financial freedom, and will value Instadapp if it is financially accessible and beneficial to them.

This also makes sense because users who are working with higher amounts of wealth will most likely be well-able to use other protocols anyway, managing funds in various places is something most users will have learnt to do anyway. This means Instadapp possibly needs less experienced users who are likely to value the convenience and safety of the product.

I think in order for the INST token to ever actually capture any value, the economics must be considered - as far as I can tell a low cost/high volume is the approach to take.

2 Likes

These two seem to make the most economical sense to me.

I would agree with this, would love some data on the Median Size of wallets on the network. This makes the most case since I believe most of the people that greatly benefit from our system are large wallets so % fees are a big turn off. A simple flat fee structure should provide us a good revenue stream while keeping us competitive.

2 Likes

The user is already likely paying the fees on their flashloan activity as a percent. .09% on AAVE / .05% on MKR so while the flat fee would make sense if the user wasn’t already likely paying on the entire position. From a UX perspective following suit with other fee models in integrated products / competitors makes more sense too IMO. I have some supporting data below.


The question I want to raise is whether or not it would be possible to charge fees anyway.

The only comparable product that I could find is Defisaver, I’ll detail their fees below and provide a link to their blogpost on fees. Most of what is used on Instadapp would fall under their “Advanced Interactions” AFAIK. (Later edit, Furucombo was mentioned on the Discord, they are currently free to use but say that they plan on charging fees that will distributed to their token holders this year)

https://help.defisaver.com/general/are-there-any-fees-for-using-de-fi-saver
Introducing COMBO Token. COMBO Tokens holders are incentivized… | by FURUCOMBO | FURUCOMBO | Medium

  • Standard interactions (e.g. Add/Withdraw collateral; Borrow/Pay back debt) have no service fees.
  • Advanced interactions (Boost, Repay, Create leveraged position, Close leveraged position) have a 0.25% service fee (of the transaction size).
  • Automation adjustments have a 0.3% service fee (of the transaction size, per automated adjustment).

Therefore a fixed attractive fee for various operations could be used, with the emphasis on reducing those fees further for smaller users, enabling Instadapp to support wealth management and distribution for all - DeFi is not just for the wealthy, the crypto community deeply value their opportunity for financial freedom, and will value Instadapp if it is financially accessible and beneficial to them.

On the note of a flat fee being democratizing/increasing access I tend to disagree because charging someone with a 100k position the same amount to do the same action as a 10m position will leave an orders of magnitude difference in the amount of fee paid as a % of the total transaction, with the benefit going to the 10m position. The flat fee structuring would have to be very robust with many tiers and could possibly be gamed if it was offering discounts to certain tx sizes/tiers.

Also, consider that many recently came to leverage farm AAVE LM rewards where Instadapp offered nearly 40% on stables through the one click strategies vs 10-15% if you just made a deposit on AAVE at that time, taking a fraction of a percent on a 300%+ increase in yield isn’t going to drive users away in my opinion and benefits all parties.


https://duneanalytics.com/queries/2995/5787
This dashboard may be depreciated, but assuming it’s not, $12b/40k (assuming there are 3,000 users on Polygon not counted on the dashboard) leaves you with an average volume per user of $300,000(roughly.) I’m sure that’s skewed higher by a few large accounts but it’s still worth looking at in the meantime.

3 Likes

Interesting data points.

A question. So DeFiSaver charges a 0.25% service fee and the flashloan charge is separate from that user has to pay or is included in this 0.25%?

1 Like

Hey Samyak, the fees are bundled. So if the action calls Aave to borrow from, the .09% fee is paid in the .25% fee from DeFi Saver.

3 Likes

Hi, thanks for the feedback and all the time invested into obtaining data and info for comparison, it’s genuinely useful.

Having read what you’ve written, it makes me remember that Instadapp can be used (to some extent) on Polygon, which does add that level of accessibility for those that really need to control costs. If the Polygon implementation were to be developed to be more comparable to the service available on Ethereum, that could be a trait strong solution, it would make Instadapp on Eth open to charging decent fees for those that are working with large funds, whilst the cost conscious users could benefit from increased opportunities on Polygon.

I’ve no doubt that would take a lot of work, but in a way it kind of makes sense as there is plenty of reward awaiting better user of polygon.

1 Like

Just saw this update from the team over at Zapper regarding adding a fee to their Zap product. Yesterday they announced there would be a 40 bps fee for providing the service but has since retracted and instead will focus on monetizing NFTs.

Seems like the Zapper community wasn’t a very big fan of the Zap fee, maybe something to consider when designing revenue model for Instadapp.

Yesterday’s Update:

Today’s Response:

4 Likes

I think, we should take fees - but it needs to be handled with grace - start them small, make sure they are clear in products, and make sure people really know what to expect ahead of time.

If we are planning to charge on any of the exisiting offerings, then let’s keep it to the lowest basis points and give a quarter or two in advance notice to users. Or in my opinion, better to just charge on new offerings (like the upcoming DeFi Limit Order). We will let community decide the next move.

6 Likes

Agree. Depending how the fees are handled, a user could offset paying fees by holding INST. I do like charging low bps on new offerings, Instadapp community is already adjusted to current strategies and functions without fees so if they do decide to use new functions in future, there wouldn’t be an alternative (without fees) to base that experience off.

4 Likes

Given the recent developments, we feel it would be best to not charge on existing offerings and to only charge on new offerings, such as the Automated DeFi Limit Order. Additionally, we would recommend keeping this fee low, around 3-5 bps, adjusted for size of transaction/upside. This is purely based on the fact that this feature isn’t available elsewhere and the majority of those who will use this feature will be the ones with larger (>1mn) wallets.
The revenue that comes from that could then be split into treasury and going back to the provider.

4 Likes

now i know what that weird flash loan looking thng was 7 hours ago on an account i hadn’t been on for the whole day… …