[Proposal] Marketing/Growth grant

Thanks for your proposal @tongnk, it’s really interesting and I agree with its importance.

Something that’s not clear to me: how are you going to deliver the milestones you mention (e.g tech documentation and SDK)?

Are you proposing to be the organizer/recruiter of this project and you can decide on how to distribute the grant further to the people delivering the individual milestones over time? Of course you as project leader would keep some of the grant to yourself, ideally subject to how successful the community thinks the project was at the end.

Would you please make a more concrete layout of expected costs/ time lines? You don’t have to only ask for BAL as it’s going to be mostly vested, if there’s need for expenses to be paid promptly Balancer Labs can also pay in DAI/USDC.

Looking forward to it!

1 Like

My skillset is in product so I usually have to partner with the necessary people - as you can imagine an SDK vs build an onboarding for LPs have vastly differing skillset and role requirements which is why it has been hard to simply put a $ figure.

The other very hard problem is that as with most things in product you don’t have a very concrete deliverable but rather an outcome that you want to achieve. A great example here is around OKRs - let’s take as an example that we commit to building out an onboarding process in 1-2 months. Let’s say in month 1 we do a very early MVP (minimal viable product) of the onboarding and deployed it. However we found that users actually need more education instead as the onboarding helps but they don’t understand the underlying. In this case because we committed to delivery the onboarding we would continue delivery of the onboarding even though the data shows that this is not the ideal case. If however we were to measure against a tangible outcome then the team is incentivised to optimise for that outcome rather than some tangible deliverable.

Another point to consider with this as well is time horizons - a very simple example of this is if we were to take a very short time horizon, (lets say 3 months) it would automatically remove some projects that we should be doing in the longer term. So things such as SEO which industry suggests takes at least 6 months of dedicated work before we get decent organic results.

However I do understand the need to keep everything accountable given I am an unknown from a quality of work perspective. What I think could work would be that we split this into 2 distinct grants, with one being a discrete piece of work that proves the quality of work and then the second one being ongoing optimisation.

1. Experimentation infrastructure and reporting:

This can be a very discrete piece of work. We can set up the infrastructure necessary for any future growth team to understand what the usage looks like. The key things that we would have to do here include:

  • Front end event based analytics
  • Consumption analytics (ideally the univision team will plug it into some form of DB so we can measure everything, given this is the key metric we need to report against)
  • Any ancillary analytics we may need to run

Ideally this grant is to allow for the community to see the quality of the work and to understand a bit more about the methodology that I am proposing. A couple of things that I would love to learn more about for point 2 below would include:

  • What does user behaviour look like for LPs
  • What does user behaviour look like for traders?
    – Do these traders use the UI or is it routed through
    – What is the MAUs look like
    – What does retention look like for users
    – What is the average value for users. Can we segment these
    – What’s the volume of automated vs manual trades
  • How often is documentation viewed
    – How often do people actually build on top of documentation
  • What is the current balancer ecosystem/landscape like (this is probably more research that I need to do regardless but will probably be useful for core team):
    – What applications are being built on top
    – Are these open/closed?

Costs + timelines depends here - I’ll just assume that I have to recruit and find someone:

  • Time:
    – Recruitment: say 2-4 weeks
    – Implementation: say 2-4 weeks
  • Cost: depends again on who I can find and their motivations so I would say either 10K USD in BAL vested, or 5K BAL vested + 10k USD
    – The other cost consideration would be costs associated to any tooling + storage. I assumed that core team would want control over this vs an outside team managing it. Happy to maintain it and just pass on costs separately.

2. Experimentation, Optimisation and Growth

I would propose that this be an ongoing grant with a couple of key metrics in place. I believe probably trading volume, MAUs and some combination of the two may be useful.

Usually how I run these is that there would be some form of product council on a quarterly basis where key stakeholders agree on the goals of the business + the metrics that we should measure these by. How the team then achieves it is up to them.

I will preface to say that these ideas will probably change drastically once we understand the behaviours from (1) but the ideas I listed before would fall under here:

  • Better documentation
  • SDKs
  • SEO
  • Content/guides
  • Functionality like onboardings etc

Cost: I would probably say 10K equivalent split between BAL/USDC here could work. If costs are unspent we could either return funds or roll them through to the next month.

Timing: we could say run this quarterly with review at the end of the quarter to either continue or halt work.

2 Likes

I think that it would be great to use your skills for such stuff.

It would be easier to vote on if there are clear deliverables and milestones, maybe it can help to separate this proposal in multiple proposals. For example: Proposal 1: doing an SDK to get BAL mining factors. Proposal 2: Improve documentation with X and Y. Also i think a proposal should be something that the applicant(s) can deliver. If you going to need to recruit someone to deliver the grant there is still IMO a too large chunk of uncertainty. A “Who” section in the proposal would help too.

Just putting in what I put into the grants channel here so people can read through:

I did outline in a response to Fernando that I believe that we can split this into 2 phases:

  • Phase 1 is a clear deliverable of analytics infrastructure to help guide where we should focus effort
  • Phase 2 is however not a clear deliverable. I’ve outlined the reasons why above but I believe the chats so far have been to measure quarterly based on key metrics with weekly reporting/catch ups/calls to update community on learnings + what has been done to date

In regards to the team, I have found 2 people I’ve worked with for phase 1 and happy to provide details if need be but phase 2 we have the same problem. The data and insights from phase 1 will ultimately lead us to understand what we should be focusing on and what resourcing we need. As a superficial example, content delivery vs a SDK has vastly differing skillsets. It’s simply not possible to understand what the make up should be now until we clearly understand what the problems are and what we need to do

1 Like

After hearing mostly positive feedback from the community, Balancer Labs decided to accept this proposal with a grant of 666 BAL (approximately 13,000 USD at the time of writing) which will be included in the first pre-approved grant batch. Infrastructure costs will be paid by with Balancer Labs own funds.

The proposed grant amount was 10,000 USD but there was a discount of approximately 25% because the BAL will be vested over a year as all BAL grants.

We’d like to invite the community to help @tongnk and also make sure he is doing a good job delivering the proposal!

Thanks so much @tongnk !

2 Likes

@tongnk - Do you have any experience with marketing / pr?

Seems like you are going to use the funds to recruit someone who does.

Yep I do product + growth so have done this before. I don’t specialise in a vertical, so for example I am not an SEO expert and I would have to bring someone in for that. But have seen achievements of double digit MoM growth in 2 previous companies through the methodology I’ve proposed

I saw another proposal for you asking for $500k - you are just all over the place.

Update #1

Quick update here.

Initial analytics is up here on Dune: https://explore.duneanalytics.com/public/dashboards/OEWehd8fuDBsEdyjuBFBs33wCK04dp1EhBhw0MUU

Other things that we are in the process of looking for are:

Insights to Date

  • Average value per trade has increased over time. A hypothesis is that this is simply a confirmation of increased liquidity means more arbitrage opportunities. Once we get the breakdown of trades done direct (via contract calls) vs exchange it should confirm/reject this hypothesis
  • Average trades per trader is lower than I thought it would be - at a peak of 14 per week. If the hypothesis of arbitrageurs dominating volume then it lends support to @5325235235235 where arbitrage is only large volumes. We should wait and see what happens with multi-hop and if this increases again

Edit 1: Fernando asked for some LP based data as well as some other items. Updated the list

1 Like

@tongnk Anyone can see these details on the dashboard. There are no “insights” here.

Seems like you were given this project without any due diligence on your capabilities.

1 Like

Hello guys, happy to assist/help with my expertise if required

Hey folks - Michael here. I head up marketing at TrustToken by day & run my own crypto-specialized marketing agency Truth Cartel by night (clients: Algorand, Compound, Republic…).

I’m not here to pitch work.

Rather, I’m writing my first post here because the process by which a not-insubstantial amount of money is being allocated to a marketing plan the equivalent of “Have you tried turning it on and off again?” seems to me overeager at best and negligent at worst - and I suspect that’s solvable by adding more marketing expertise to the conversation. So I’m here to try to help that, though we’ll see how effectively.

If it’s helpful, I’d love to have this discussion (inspired by our investigation into cohort analysis) as a general thrust towards developing some substantive strategy - ie, what do we need to say to whom and where:

  1. Who: Archetypally, who are the folks who would most benefit from what Balancer has to offer? Day traders? Yield farmers? Institutions? Good trick is to start with yourself as an archetype. This all tells us who we’re speaking to.
  2. Why: Why is each archetype coming to Balancer? What’s their primary benefit? This tells us what to offer these folks to attract their business.
  3. Where: Where did they/do they learn about the benefits of what Balancer has to offer? This allows us to reverse-engineer where to reach them.
  4. How: How might these new users be dissuaded from using the platform? This informs us about how we might reduce friction (ie improve UI/onboarding), improve education (struggling with getting a token pair listed, need to guide them through governance process), or learn from competition (they’d go to X competitor instead because Y)

This isn’t some famous marketing model. It’s just a destillation of what marketing really is whether for Balancer or Nike: conversations with individual people - at scale.

Oh, and I also volunteer to audit the resulting proposal, if that’d help.

2 Likes

This is a good post of what I’ve been talking about both in governance and here. The idea was to start with the data to build out the who - if you have a look at the Dune reports we can see people who interact directly with the contract vs those who interact via the UI.

However given the resistance to analytics from the team around privacy we have actually shifted the focus slightly towards:

  1. High level analytics dashboard
  2. Build out guides for 3 personas (traders, LPs, devs)
  3. Build out more content w.r.t. how Balancer works (move away from technical + maths jargon)
  4. Build out functionality and tooling for LPs + traders

More than happy to chat (maybe on Discord?) - I think there are definitely synergies in skillsets. My background is in product so growth has always been focused around experimentation, data, and feature optimisation.

The idea was always to decide where we go once we answer some of the key questions listed above (e.g. if there is no retail traders and volume is all in direct contract interactions then it may not make sense to focus on retail traders given the strong presence of Uniswap already)

1 Like

Posting here just to keep everyone updated. We realised that Dune Analytics unfortunately doesn’t have accurate pricing data as they use CoinPaprika instead of Coingecko. So TVL shows ~350m instead of ~600+m

So instead what we agreed on was to build out a dashboard plugged into subgraph to display the following:

Single numbers:

  • TVL
  • Total trading volume
  • Total fees earned
  • Number of public pools
  • Number of private pools

Historical charts for:

  • TVL
  • Trading fees earned
  • Trading volume
  • Liquidity utilisation (trading volume / avg liquidity over a 24hr period)
  • Revenue ratio (fees / avg liquidity over a 24hr period)

The key ones we should be keeping our eye on is going to be that liquidity utilisation and revenue ratios as the end goal is for profitable pools without a BAL reward

1 Like

@tongnk hey just a kind word to say that many silent readers do support your effort and the updates you provide. Looking forward to see what you will deliver.

4 Likes

Just wanted to post a quick update here and get further thoughts as well.

Quick Update

It’s taken us a bit longer than expected but I think we’re nearly there. Just a quick sneak peak before we launch. We are going to have 3 sections currently:

  1. General stats
  2. Past 24 hour stats
  3. In depth chart stats

General Stats

Are these cards. All of them are planned to be cumulative stats. Some notes:

  • We’ve got a little tool tip to explain any metrics that have calculations
  • Graph inside is a 30 day window
  • Percentage is the change from 30 days ago until now

Past 24 hours Stats

We added in some stats here. The two important calculation ones are:

  • Liquidity utilisation: trading volume over past 24 hours / TVL
  • Revenue ratio: fees earned over past 24 hours / TVL

Ideally these two ratios should tell us how we are tracking over time. With the goal of pools being self sufficient and IL not being an issue, we need to see these %s rising (especially the revenue ratio one)

In depth chart stats

Still playing around but this is how it’s looking so far for total locked value. We have the two components here:

  1. Total Value Locked: this is a cumulative number
  2. Total Value Locked volume: this is a daily change. Red = negative, green = positive

We’d love some feedback!

We’d love to ask for feedback!

  • Are the graphs intuitive? Do they make sense to you?
  • Are there any key statistics that are missing?
  • In the volume in depth chart does it make sense to have red/green or would it make more sense to have a bar going up or down around 0?
  • If you notice in the in depth charts there is a drop down arrow which is for toggling into trading volume and fees. Does it make sense to put it here or do you want to see them all as separate charts?

Things to Do

These are the following things we are working on before the completion of this grant:

  • Liquidity ratio over time (see explanation above)
  • Revenue ratio over time (see explanation above)
  • Daily trading volume (we have the cumulative part but need the daily trading volume part)
  • Daily fees earned (we have cumulative part but need the daily part)

Question for the future

We have ideas around extending these analytics to provide some more in depth features under a new grant but would be interested to see if there is any interest in any of the following:

  • Top performing tokens: list of top 25 tokens along with their liquidity and other key statistics
  • Pool performance: a new page for each live pool to understand a couple of metrics over time
    – Volume traded
    – Fees earned
    – BPT price
1 Like

Phase 2 Proposal

Thanks everyone for the positive feedback in the discord channel - it seems we definitely want some better analytics as a community! I’m just going to post in here instead of the discord so it doesn’t get lost. For the given proposal I’ve broken everything down so that the community can decide what they would want to built and what they don’t want.

Just a quick note: What I’ve left out here is anything to do with specific wallet balances - after talking to 1saf and doing some further digging there seems to be a lot of work to get this working in a way that we would be proud and happy of. We believe that the following should be very good baseline that is required first before we explore wallet balances.

Tokens

We would want to have an area to display basic token statistics in a table form that is paginated. The concept behind this is to allow people to discover which tokens are driving liquidity. We would look to include

  • Liquidity
  • Price
  • Symbol

There would also be a search functionality to be built and potentially some form of filtering functionality to allow for token narrowing

Grant Amount: $2,000 USDC

Pools

This is where the main body of work would be done.

Dash

The concept is to display a main table of all pools with historical performance that then allows them to click into a specific pool page. On this view we would look at displaying the following 30 day historical stats

  • Name
  • Tokens + weights
  • Liquidity
  • Trading volume
  • Fee volume
  • Returns (take the BPT price and the difference between now vs 30 days ago)

In addition we would have some search/filter functionality

Grant Amount: $3,000 USDC

Pool

This is where we’d display key information about each pool. We aim to split this into the same as the main dashboard.

We would in addition put links for:

  • pools.vision to display BAL rewards
  • Balancer pool to add liquidity
  • Link to exchange to trade

Single Numbers

  • TVL
  • Total swap volume
  • Total fees
  • Revenue Ratio
  • Liquidity utilisation
  • BPT price

Graphs

  • TVL
  • Total swap volume
  • Total fees
  • Revenue Ratio
  • Liquidity utilisation
  • BPT price

Grant Amount: $10,000 USDC

Note we aim to have these all mobile responsive as is the current dashboard

Total Grant: $15,000 USD

Thanks for this proposal Nick, I think it’s great.

The one addition I would make is to include also the comparison of the BPT value with the value of holding tokens (AKA impermanent loss). So instead of saying your BPT is worth 30% more than a month ago, you can say your BPT is also worth 10% more than your tokens would had you held them separately. This is what uniswapRoi does really well.

If no one raises objections Balancer Labs would like to move ahead with this, especially since the grant you are asking is actually going to be paid in USDC as a company (Balancer Labs) expense and not as a BAL grant from the ecosystem fund.

Yep this should be doable. So basically we are trying to reflect 2 portfolios starting at the same value and seeing what the divergence looks like.

I think though we would have to use Coingecko instead of the internal pool pricing (sometimes subgraph doesn’t seem 100% correct for token prices and I’m unsure how the underlying works) so if there is no Coingecko pricing then we wouldn’t be able to display the HODL scenario.

We’ll finish phase 1 and we’ll start scoping this out in more depth and get to work on it now then :slight_smile: