Growth Strategy

I’m writing this post to combine several growth related topics into one high level strategy.

I think growth means volume on the perps side and TVL on the lending side. I see it happening in two steps. First we make the product very compelling. Then we use the tools available at the DAO’s disposal to ramp up marketing.

Tasks

  • proper stats on pnl, interest earned, funding earned, deposits, withdrawals, leaderboard
  • stop loss (maybe other advanced order types)
  • 10+ perp markets with liquidity equal to other DeFi perp protocols (e.g. $2m size in BTC-PERP within ±20bps of mid)
  • Plenty of RPC nodes and support servers
  • liquidator that doubles as a market maker

Tools

  • Set taker fees to 0bps
  • offer volume based fee tiers/rewards (if possible in smart contract)
  • offer MNGO based fee tiers (if possible)

Once we have the compelling product, which I think can be done in a month, we can start using all the tools at the DAO’s disposal to bring in more users. The main tool would be setting taker fees to 0bps, at least temporarily. And of course we can all go out and tell people about Mango vigorously. The reason I prefer to stabilize the protocol first is I want anyone we attract with fees or rewards to be sticky and stay in Mango for the superior experience. Right now, Mango doesn’t really offer a superior experience. But I’m confident Mango is the best protocol in its foundations.

I’m inclined towards the 0bps fees and against the tier based advantages because of my Simplicity Uber Alles approach, but I think there are strong arguments to be made for the other approaches and I’m willing to switch to those if the Simplicity way isn’t working. The main reason to prefer simplicity is it sets us apart from other protocols/platforms and is quite egalitarian. Everyone gets the same fees and there are no tricks or special formulas to understand. I’m super hyped to promote this message far and wide. But obviously we can’t stick to a losing strategy. If it doesn’t work we chuck it.

2 Likes

My order of importance:

  • The RPC node thing is such a pain. It is the #1 source of user frictions (both on UI and programmatic). This is a major threat to creating great experiences for users and risks actually giving such awful experiences that people are permanently turned off. When it comes to trading, especially on high leverage, it is unacceptable to have delays on the order of 10’s of seconds, confusing / wrong error messages saying something has timed out when it actually went through. This risks traders who think they’re making a 1 BTC long position taking a 5 BTC long position because they are trying multiple times. This problem could destroy the Mango Markets if it is not taken seriously.

  • There needs to be stop orders, and reliable ones. Traders who are using leverage like this want to control risk, go out to dinner, set a stop loss, and reliably know that if the market goes against them, their risk is adequately managed. There are a class of traders for which the current product with no stops is a dealbreaker.

  • Proper, downloadable and full account history and exchange-level data of interest is so important. Doing the accounting on Mango is a nightmare. How much did I earn in interest? How much funding have I paid? Was my liquidation fair? None of these questions are easy to answer. CEXes struggle to provide this stuff as is, so idk how it gets solved on Mango.

  • Taker fees at 0 bps is just shady. 0-fee exchanges = always a red flag in this space. It needs to be at least 1bp to avoid this perception that the volume is utter garbage and Open Interest as a measure is just not a substitute. However, the current market, in current form, is too sticky in the prices. Need to let traders capture prices and get in and out vs existing liquidity.

  • Volume-based fee incentives and/or MNGO holding incentives are a good idea too IMO. I don’t think having this in place is gimmicky or tricking users. Markets are not egalitarian. A minority of users do a vast majority of the volume – and those traders are more important and deserve to get lower fees and other benefits.

9 Likes
  • The Solana devs are working on fixing the RPC nodes and websockets. We’re also working on improving the reliability here. But yeah, I agree connectivity should be extremely good.
  • Agreed. Stop orders are incoming. My only hesitation is whether I should build it myself or try to grow the number of devs by getting others to build it. Either way, this will have to get done soon.
  • In the short run, we’ll just gather the data and provide it, but I think when we have the time, the DAO should start compensating for providing this. Either that or we have to think of business models on this. It’s a lot of work to get high quality data and it’s not very easy on Solana.
  • Ok you’ve convinced me 0bps just has a bad reputation. So perhaps the promotional rate can be 2bps.
  • Agreed markets are not egalitarian. But not sure if a few traders do most of the taker volume actually. Imagine if Mango got all the people with smaller books and people with bigger books stayed at other venues. I think that’s a good scenario. I’m inclined to think the big books follow the small books than the other way around.

FTX has shown that first attracting people who trade with size is a winning strategy for a new exchange to pull off. It’ll be a lot easier to convince two-dozen knowledgable whales to use a better product than an army of traders with $10k accounts. MNGO holding or volume incentives should be very low priority, it won’t have much of an impact at all until base fees are higher and/or negative maker fees are possible. So I agree with the urgency of @daffy’s “tasks” and don’t think the “tools” are necessary.

Agreed on the stickyness being the most important KPI. In terms of steps to improve it, I agree with swap:

  1. Latency & Timeout issues
  2. Stop Loss
  3. Interest Accounting

All three need substantial investments in off-chain infrastructure or really big changes to the on-chain program architecture. Maybe worth a band-aid in server bill.

10+ perp markets with liquidity equal to other DeFi perp protocols (e.g. $2m size in BTC-PERP within ±20bps of mid)

What are the different challenges to achieve decent liquidity for all kinds of markets?

Also would mango markets be interested in listing markets which are short term trending? Many such markets are immediately available (which attracts user to these venues) on ftx/binance but slowly peter out to zero over time

We only have place for 16 tokens and don’t have code to unlist markets rn. So i’d say we should be more careful about listing assets that won’t show strong volume for 3-6 months.

Liquidity and Volume are usually strongly correlated. As long as there are takers, makers will be able to make a profit providing liquidity.

3 Likes

All good points.
Just want to step back and analyze this.

What is the goal, from mango project pov?

  1. high trading volume, with “good” trading experience.
  2. “good” means quality liquidity, smooth execution, and other user experience.

To achieve the goal, I would prioritize like this:

  1. Infrastructure work like rpc etc. I get a lot of errors/delays/misleading info while trying higher freq mm.
  2. Incentives. Try balance out retail takers/makers, whale takers/makers, quotes, and trading volumes. Even among pros, there are small whales(is there such a thing?) and big whales. so exp(8) may be too extreme. Each one needs to be taken care of and not totally ignored. And each one, favored alone, brings issues. For example, a lot of exchanges reward by trading volume, which can encourage self trading. But self trading has its own risks so reward by trading volume may not be as bad as you think. And we all know the dynamics of the current reward structure.
    The reason every player needs to be taken care of is, if people feel after they spend a lot of capital and time, and it is still a losing game for them, they will leave. I know it is easier said than done, but still.
  3. Taker/maker fee table. It is ok as of now, imho. It is not thin margin/heaving trading yet, so fees are not important to people now.

I may even suggest try all different options as pilot programs. Only actual deployment will tell which one is best.
While $1.7m 24 hour volume is not bad, it can be better. Need more takers. Yet another reason to reward by trading volume.

Not to beat a dead horse, but when I see 0 bps taker fees I’m instantly suspicious of the platform. It would have to be very clearly advertised as a temporary measure, maybe with a countdown or something, to make it clear it’s operating at a loss. Even then, as others have mentioned, it would attract shady characters and shady practices. And as Austerity mentioned, it’s a blow to exchange reputation. We would be much better off creating some attractive features and fixing the myriad of issues we’ve all pointed out first, then bringing down taker fees if we’re stagnant. Give the platform a chance to grow organically.

Would it make sense to flesh out docs about how to setup a dedicated RPC node? Users would feel less helpless knowing there is some other option while the delays are getting fixed in the pool nodes.

2 Likes

A few thoughts about advanced order types:
An integration of order types not currently available in the big exchanges could attract some people. Now these order types might not be the easiest to implement though:

  1. OCO orders (one cancels the other) → You can have a buy and a sell open and when one of them triggers the other one is cancelled
  2. I dont really know if it has a name. It would be nice if you could set a percentage and when the highest bid comes close to your sell order and passes the defined percentage gap, it would cancel (and vice versa). Now this somehow conflicts with the market making incentives and it invites spoofing. But I think its an attractive tool to play with. And the liquidity would still be there for someone to swipe in one order.
  3. The option to define a maximum price for a market order to define the maximum slippage. It fills up to the defined price and deletes itself after

EDIT2:
4. An option to create an order ladder within one transaction on the chain. You define a percentage gap between orders and the number of orders, the starting price and the total size

3 Likes

I love the idea of having complex/advanced order options and/or UI tools that aren’t available elsewhere. If we can pull it off then there won’t be as much pressure on listings, since our trading mechanics will be completely original and create an appealing draw for traders (complexity allows convenience).

Focusing on listings would maybe attract some volume from random crowds, but focusing on UI tools and exotic mechanics would target a more serious trading demographic, which is ultimately what we want.

Even something as simple as a position calculator that provides slightly more insight than any other exchange calculator around would be enough. Apprehensive traders may start using the calculator while they trade elsewhere, and eventually trust the mango project enough to start trading on it.

Everyone has listings and continues to list, it’s exotic tools that will be the real draw.

We could go a step further and make certain tools available only to mango holders. Let the token directly represent the value of these tools by acting as a gatekeeper.

3 Likes

I think emphasising the fact that mango is a borrow/lend platform would attract more TVL inside mango and allow for larger traders to join. Seeing how the newer platforms providing this service have already crept up or overtaken mango in TVL.

Right now the liquidity mining program only really appeals to sophisticated users (generally makers). To increase volume we need to find ways to appeal to less sophisticated users (i.e., takers).

To do that, I think the most common thing for cexes is referral links. The issue for a dex is preventing people from abusing the ref link system (i.e., using their own ref link).

There are some ways to do this, like tiering rewards by volume, so that even if you do abuse the system you aren’t taking more than you are giving.

I also think it might be cool to implement some sort of NFT badge system, for X volume or number of trades you get certain awards.

1 Like

Difference is really that all “lending protocols” are using token emission to incentivize lending & borrowing with beefed up APRs in a struggle to optimize for TVL. TVL really doesn’t matter, our core KPI should be daily perpetual volume.

4 Likes