We voted against this, but for different reasons than other delegates. While we appreciate the general direction of this proposal and think it could work given a narrow enough scope, we don’t think the design works for a larger ecosystem with many kinds of projects in it.
The biggest problem is that the proposal assumes a standardized approach to CAC across DeFi projects, which doesn’t align with the reality of how conversions are defined and measured in the industry. In DeFi, conversions are bespoke and vary widely depending on the type of the app and its goals, with every project using a different method across a spectrum for what classifies as a conversion.
For Vertex, we qualify a full conversion as someone that fills a trade order, reflected by the event it emits, and is trackable both internally and via Spindl who is integrated with our SDK. Comparatively, many projects simply use deposits as qualifying for a conversion, and ad publishers such as Coinbase Wallet use a significantly lower bar for conversions, which is just whether or not someone clicked on the link and landed on the app domain.
If we pay for ad clicks that don’t result in trades but are counted as conversions by a publisher, our CAC appears artificially low, while our actual cost per meaningful user is much higher. This also becomes easily manipulated by projects using low thresholds for conversions. When you scale this problem out to multiple projects, you end up with distorted data, painting a inaccurate picture of the success of each campaign.
The primary way we see to fix this would be to set a universal conversion methodology for every participant, but that would again require an extremely narrow focus as we don’t think there’s a methodology that could be widely applied here. If conversions aren’t aligned with the stated KPI outcomes like TVL / maximizing trade volume, the program risks misallocating its budget on shallow user interactions.
Separately, we also have concerns around the application process itself. While we don’t mind sharing some information, we’re not sure that we would want to reveal all details of marketing we’ve done in the past year, as this quickly ends up in areas where one strategy is a competitive advantage over another, and in the end projects are still trying to draw the most users to themselves.