Data-Driven Tokenomics: Analyzing Incentives in Action by Lisa JY Tan | Devcon SEA

[Music] have you guys had a good lunch yeah okay fantastic we only have five minutes and then we have two minutes for questions so I'm going to go really really fast going to pay a lot of attention for five minutes so today we're going to talk a lot about understanding econom uh data and how data impacts economics who here build economics or who here deals with economics okay who here are founders it cares about economics and who here are Builders great good wow a very very even split so just a tldr of of economics and data when people think about economics people usually think about it in a zero to one phase when I need to launch a token when I need to launch my project I need to think about tokens and how tokens interact in the space and that's where a lot of the focus has been which rightly so things like your modeling your your agent based simulations your risk testing your risk scenario analysis that's a very very big portion at the same time today I want to I want you to start thinking about how can we use data to start understanding how the protocol has been doing and use data to help improve the health of the protocol in every project there are two key phases 0 to one and one to 100 0 to one is when you're brand new you're fresh is MVP finding product Market fit going to Market and there you use a lot of of economics and to build it's all Theory based to build out the economics on the other hand when it's live in your one to 100 you actually have user feedback dat live data validated data that you could stress test the system and really understand how users behave so the challenge that I want you guys to start thinking about is how how can you how is the data giving us information about how users are interacting with the protocol the truth is that as much as you're building the the economy you're defining the economy you're writing out the rules of how this game works and people are just going to game the system whether you like it or not people are just going to game the system they're going to find post they're going to find ways to extract more API more value and that's fine that's normal in fact once you launch the project things are just going to get more and more chaotic if everything follows the model and everything follows the simulations that we built out all the economists that I know they be really really rich but they they really economists are quite poor people so the reality is that it's not like models it's not like simulation so you need to be able to balance that and one way to start thinking about it is using a lot of sess metrics sess analytics things like a average revenue per user user acquisition user retention all these data will be very valuable and insightful to understand how users are behaving and interacting with the protocol any of the day tokens are just an incentive to to get people to behave in a certain way so once we have the data we can validate if users are really reacting to these incentives or not and it's fine if they're reacting to it double down if they're not reacting to it how do we change how do we update the rules update the system so that is the that that's the general gist happy to go into more case studies happy to answer more questions and I think there'll be time for questions do you guys want to raise your hand and see how far I can throw throw this damn thing okay I'm going to throw it and you try catching okay hey good job all right go ahead um hey everyone well actually you can talk I talk in this oh that is crazy did not I thought that was just a fun exercise we were doing um exercise go to k fit yes I love your videos I love just everything you do um from your perspective what has been the most like successful toonomic design in terms of adoption from a data perspective um so I know you know for context you see a lot of um uh tokens you know or just protocols use LP incentives right but you know if you look at the data um that you know Unis swap has put out they typically aren't very sustaining even though they might you know cause like temporal um advancement in terms of what you're trying to achieve and it seems as though there are not too many sustaining toonomic designs that actually drive you know you know wanted Behavior over the long term so just wanted to know your thoughts on that yes very good question and as an economist the answer is it depends however there's context to this it depends on two main factors your business model and your Revenue stream if your main goal is to get more liquidity in there and then you you need these incentives to boost strap liquidity versus you need incentives to keep getting more people to to interact with the protocol these are all different kind of behaviors at the end of the day the question you're asking is what is the best incentive out there my question to you then is what is the behavior we're trying to incentivize are we trying to incentivize user acquisition or user retention take air drop for example which is a very popular thing which I not the biggest fan of but it's a great distribution mechanism the pro the pros of that let's start with Pros huge user acquisition however the con of that is there is lack of user retention people are there just for the financial gains and as I said when when you put the rules of the system out the game the rules of the game people are just here to exploit the system so which is fine maybe you could start off with with user user acquisition and some air drops but it doesn't need to run for such a long time take the data and understand understand how users are are reacting to it alternatively it's point system where youve heard a lot about and also how do you allocate reputation to different kind of people so that they can and then we can start allocating it correctly to the right kind of users hi so uh thank you great talk and uh um my question is more around the social aspects of a platform right so H toomics that it's easy to understand not easy but I mean it's different to understand uh rois and things that are more financial and and like onedimensional but on social like I'm from graa XYZ and we're trying to to optimize social interactions and meaningful social interactions I wonder if you know some examples of where this comes in and where you saw like remarkable examples of like toomics that work with the social aspect that's a very good question because everything that we have been talking about so far is always the quantitative because quantitative is something that we can measure your apys your tvls all these are something we can measure and once you can measure that you can always optimize for that and I really love your perspective which is social social is really hard to quantify it's really hard to Define and one example will be things like governance or decision-making protocols so one thing to think about or one one Avenue to start considering is how can we tie in the the natural way that people go inclined to which is financial incentives tie it with the social aspect take poly market for example right the biggest case study of of this year how do you use Financial incentives as a way to Signal preferences and from there you can start building very very complex models and very complex decision-making mechanisms that takes into consider consideration both the social aspects and the qualitative aspects and the quantitative aspects happy to go more into details later oh yeah we have another question over there thank you thank you very much so I wanted to follow up to the first question because you said that different type of activities require different type of incentives which is uh I totally agree with that so we have analyzed for example LTI data from arbitrum uh where arbitrum spent $15 million on incentivizing Defi and apps and user acquisition for them and the thing is that most of those uh users basically left at DBS and defies after three months uh either to move to another program for example ZK sync ignite which is pretty similar and is happening right now I mean it's work in progress so basically we call these users mercenary users so what are your ideas and thoughts about keeping the um renting these users keeping their retention the ecosystem and fighting off also mercenary users who basically move from program to program to uh utilize it for their own uh profit so I'll give you the short answer first and happy to talk about the long answer after the short answer is that these things are something we cannot control this is just natural forces of how the market works when there is green grass the Sheep would just go there and eat the damn grass that's just how reality works there's no point controlling that it's not possible to control what we can do is to understand that it's going to exist and how do we how do we channel the the funds or the assets the resources to much better highly better resources so reputation building understanding like transaction history are these people mercenary users or are they actually value creators any of the day you you are balancing two things one is value extractor in the short term and one is value Creator in the long run and what are the different nuances within your business within the protocol within within the foundation that help to channel towards the other side of things so there is an aspect of token there's also an aspect of business model and Reven Revenue stream that you have to balance quite a bit that was the last question thank you very much for your uh talk please give a nice round of applause to Lisa