DePIN Token Economics Deep Dive

Transcript

Hello everybody. I'm thrilled to be here. My name is Tom Trowbridge. I'm here to speak about deep in token economics. Um It's great to be ECC at the ninth ECC. This is the sixth one I think that I've attended and thrilled to be back here in Khan and looking forward to to more here. So what a terrific place to do this. Um this talk comes from the deep in token economics report that I published last year. And for those of you who want to understand more about token economics, I highly recommend it. Obviously, I'm biased, but most reports I've seen on this talk about token economics in theory. This looks at all the tokens or most of the tokens that are trading in the deep in space and dives into detail

on what seems to work and what doesn't. And so it's a really practical guide to token economics. Um and it needs to be updated every, you know, so often and there'll be another one coming. But there's clearly challenges with doing this, right? When you're doing a token economics report, there is a whole host of challenges. There's no standard format for describing token economics. There's not even a source. Projects list their token token economics on X, on blogs, on white papers. There's a wide range of even of detail that's shared.

And so and token models, of course, are evolving and get updated. So this is kind of a snapshot in time of what the deep in token economics looked like at that point. And I'm welcome feedback, corrections, additions, and I will publish another version of this, an updated version, later this year. But basically, the one takeaway for projects and investors is token economics should fit on one page. Right? There's no reason for it to be more complex than that. You have to have a couple key points like total supply, how the rewards work, what inflation is and how it's generated, staking, how that works, and how the revenue moves to the token.

And obviously, very important, the governance mechanism for altering these economics. And the reason why I think this is important is investors value transparency and predictability. And when you have something that's arbitrary and not transparent, you will be valued less. And I will show an example of this going forward, but I think the opacity of our space and the wide variety of how this things these token economics are structured and even described does a real disservice. And we can zoom out well beyond DePIN, right? This is true across the whole sector. And so DePIN is just taking very specific frame, but you

can look much more broadly than than just DePIN. But let's take a quick second and what is DePIN so we're on the same page. So, DePIN is crowdsourced physical infrastructure put together in a network providing a useful service that's secured and incentivized by crypto economics. All right. And so this obviously excludes mining. Um and if it did include mining, then pretty much everything would be DePIN, so it kind of has to exclude it. Um but that's what DePIN is. And there two main kinds of DePIN. There's physi- there's cloud DePINs, which are think of them as the server-based services that are the Amazon or the um um the Google or the um you know, cloud

alternatives that do file storage, they do um file uh rendering, um like you see video rendering like render and Livepeer, and they also do compute, whether it's CPU or GPU compute. Those are really the digital cloud DePINs. The other side are physical. And the physical ones usually involve a um use-specific device like an antenna for a um for something like a weather XM or Geonet or they use very specific applications on other hardware for example on your phone. Some of them do that. Layer drone does it on drones. So there's a wide variety. Dimo has a

device you plug into your cars dashboard. Right, so a lot of different physical devices that collect data and allow services to be offered and I think those are just an important frame to set what DePIN is. DePIN has been out of favor recently. A lot of crypto has obviously, but I think it's going to return because DePIN is actually a core enabler of AI and it's across the AI spectrum. And so you have model construction and training from Genshin and BitTensor and others. These are just some examples. There's plenty more. There's GPU marketplaces. Fluence, which I'm a co-founder of is a GPU marketplace. Akash has been around for very long time. There are GPU and CPU marketplace. IoTeX,

Aether and others. Then there's web data. Grass scrapes terabytes of data a day that are used in LLMs. Location. Geonet gets 1 mm specific location. Perfect for robotics. Very useful in AI. And then visual data. Layer drone uses drones to map the world in 3D far more precisely and more updated than than satellites. So these are all core AI enabling technologies that are superior and cheaper than the centralized alternatives.

So why is DePIN interesting? Just let's also take a step back for a second. Its global crypto market is obviously it's down from its peaks, but still about 2.4 trillion. So it's huge. Hardware features and pricing are only going down. Like hardware pricing goes down. Features are are up. What's available on your phone now was unthinkable a decade ago and a tremendous evolution from even 5 years ago from the optics to all the sensors. And then software availability is obviously increasing dramatically. The software sophistication and the low cost and availability of open source has also increased dramatically. So DIM takes advantage of these three trends, each one of which is moving only in one

direction. Crypto is only increasing its adoption. Hardware features are only increasing, pricing is going down, and software availability is only going up, right? So, DIM takes advantage of all three trends, which are going in only in one direction. And what are some examples of the value proposition that DIM is able to extract from those three trends? There's a wide variety. Here's a couple of them. I mentioned Dimo, plugs in your car, and now Dimo is collecting car data and offering services across models and manufacturers that was never before possible. GeoNet has a 1 mm specific location, which is 90% cheaper and 100 times more precise than most of its competitors. Helium

offers mobile much cheaper. Layer Drone maps cities by drone that ends up being about 50% cheaper, maybe even more, but most importantly, three or four times better resolution, and it can be mapped, you know, the next day, right? So, 10 to 20 times faster than a satellite or even a a plane, and it can be done in all weather or at least cloud cover. So, massively superior offering. Hivemapper mapping in cars, I'm sure you're familiar with that. And Grass is obviously offering um web scrape data much cheaper than the centralized alternatives. And so, why do we care about token economics? Those are great services. Why do we care about the

token economics? Well, it's because we're past the meme stage in crypto, right? So, meme stage is is we moved beyond that and traditional investors are moving into the space. There's now about 300 billion dollars of capital between from traditional investors from either the ETFs or the dots. Right, that is maybe 15% or so of the crypto ecosystem that also is growing. That we should and we're going to welcome that investor base, but that base cares about revenue and results and revenue they only care about if there is a link to the token, which is where token economics comes in.

Right, so it doesn't matter if you generate a lot of revenue if there isn't a link and that into the token. Otherwise, it's effectively a meme and I think we know most of crypto are memes and that is part of the reason I think we're in a little bit of a difficult period right now, but this is changing as projects generate revenue. When you're looking at token economics, the first thing you want to start with is what's the reward schedule, right? Because this is what generates the um the kind of token supply overall. So and there is a wide variety of reward schedules out there. There is no one um model, right? Some have a constant decline of rewards like Arweave, others

have a very specific 15% annual decline in perpetuity. Others have two-year halvings or like GeoNet does that um as does as does Helium kind of referencing the Bitcoin model. Others have a percentage of revenue, which relates to the rewards, which is also you know makes it makes a lot of sense um and others are more network-based and some are just decided annually. And the proposition I have is the more um fixed that number is and the more definable it is, the more investors will find it um interesting and pay attention to it. If it is arbitrary, you will suffer a discount as a result because investors

value certainty and predictability. And that leads to the challenges. Why do some of these um token rewards are they sort of open-ended? It's because they're not priced in fiat. And so people rewarded in an arbitrary number of tokens, that can lead to incredibly high dollar value rewards. And we've seen projects benefit or or have a challenge where their providers were paid so much that they didn't actually want to offer services cuz they were just getting paid in tokens whether they were service on them or not. Um and then with low token price happens, your providers who have fiat-based costs go out of business. And I think, you know, Filecoin is a real example of this where Filecoin miners

did incredibly well in the Filecoin token boom, didn't even care so much about serving real customers. And then in the price um fall, had to go out of business cuz they couldn't they fell so far they couldn't cover their costs. So, the answer here is you really need to have um rewards that are fiat-linked so that there isn't an arbitrary token amount that can be worth a huge amount or very little. That's not a way you can build a sustainable business. And DePIN projects are looking for institutional adoption, have real institutional customers. And so the people that are providing that service must be dependable, or else you're not going to get the the actual

um the actual uh customer base that you need. And you also should be pricing the service, obviously, in a fiat price. You can take tokens, but no one is going to try to budget in some arbitrary token over some multi-year time frame or even multi-month time frame because it's just not how companies budget. So, this link to fiat, as much as some of us may not like that and think it's different than than how the industry should work, is really it's it's it is a kind of a gating item for success and for growth and sustainability in the space.

Um And so why why does this make DePIN different? Well, if you're following those those comments and you are have the appropriate um reward schedule and the ability to attract institutional customers, your revenue and you're offering a valuable service revenue growth. And we're going to look at Geodnet and Grass, which are two kind of leading DePIN projects, and show kind of their profiles and how they've been doing and what they are in case you don't know them. Um Geodnet has is RTK location provider.

So, it provides locations up to 1 cm. And it is that's like 1 m versus, you know, that's what GPS is. So, this is 1 cm, right? So, it's perfect for farming, for robotics, for all kinds of things. They offer, right the last time I checked, 13,000 sites for this RTK um location service. The nearest centralized competitor has less than half of that. And that centralized competitor took 10 or 15 years to get there. Geodnet did it in like three or four, right? So, it has double the amount of coverage of its centralized competitor, is 30 to 80% cheaper, that depends on how long you buy the service for etc.

Um and it's a big, you know, it's a three plus billion dollar market. I think robotics are going to grow that even even more substantially. Um they generated 8 and 1/2 million of ARR as of March this year. All right, this has doubled from a year ago. So, they're doubling revenue over the last couple years, which is which is very exciting. We're going to see how that impacts the token in a second. Um but but, you know, farming and robotics customers is pretty interesting. Now, Grass basically you is a web scraping protocol. So, you download a browser plugin and it scrapes uses your um your IP address to scrape data, which it then collects and sells to to LLMs. All right, 8 million active monthly users

are scraping data for Grass, right? That is a massive scale. And there are no centralized companies that provided on the scale. It's it's it's doing something from um, I think 80 to 1,700 terabytes a day of web data scraping, right? Massive amounts of data. And you guys all know, LLMs are voracious consumers of data, and current data is valuable. So, this is a pretty fascinating um, business. Its revenue is 33 million as of Q1, right? So, massive um, amount of revenue for this business, even though if you're not an exchange, I guess I would say. But, here's the interesting thing. We've got these two

projects, um, Grass and GeoNet. Both, you know, GeoNet has 1/4 of the ARR. Their circulating supply is about the same in terms of value. And that's because GeoNet does an 80% buyback. So, it takes 80% of its revenue, buys and burns the token transparently on chain. Grass buys and burns, but like 5% and it's somewhat arbitrary. And so, even though one is generating 4x the revenue, circulating supply value, about the same. And that just and even though the fully diluted of Grass is is about double GeoNet, it's nowhere near where it would be based on just the revenue numbers, right? And so, I think this is just a

little indication that token economics really matter because you've got a project during quarter the revenue with roughly comparable size, which I think is due to the transparency and faith that an 80% buy burn does. And so, buy burn is used by a large number of um, of projects, Helium, Glow, etc. GeoNet bought a million dollars of its token back in Q4 of '25 as an example, but these projects use it. Here's a buy burn percentage. Globiance and burns 100% of its revenue. You know, Render buys a lot. GeoNet does as well, and then you can keep going down the spectrum. And so, token economics um link revenue with traction. Now,

marketplace ones can't do buy burn cuz if you run a marketplace, you're basically giving money from customers to the suppliers. You're sitting in the middle taking a little fee. Much harder to to do buy burn at scale. So, you but you need stake and rewards. And so, for these um cloud DPNs, you stake is used as a way to generate trust. So, when you got these third-party servers and you can't you don't know who the provider is, people or they stake on those servers, and that's how you know that you can trust them cuz they'll get burned or won't be able to earn revenue if they do not perform. And so, as the project scales, the stake required scales, and that's the is the basically token scarcity driver, um which which which which which will

see. Um one important point is you should allow not just the providers to stake, but anybody else. In early-stage DePIN projects, required their providers to stake, and that was it. That's not optimal. Raises the cost for the provider. They have to buy the token to stake it. You should have investors be able to stake on the providers. Those are often very different universes. So, it makes sense to um to have to have to basically have an ecosystem that includes both. And so, as example, Fluence is case study here. We are a GPU platform.

Um institutional users connect for low-cost GPU usage, and payments via stablecoin. There's a small fee to Fluence, and trust is via staking. Um and so, that that sounds um sounds sounds attractive. How's that play in token economics? Well, revenue is passed to stakers. And so, as example, if we get 10 million of revenue, that's about 500 That'd be about 500,000 of revenue to Fluence. That's about 1 cent to the amount of tokens we have staked today. Right? Our token price is is like half of a penny. So, if we did that, it'd be about 130 150% yield. And I only reason I mention this 10 million number is that I'm pretty confident that we're going to get

that in the near future. Now, not all of that will be passed on to token holders. We have um opex and and to to cover as well. But, that gives you a sense as to how these models can scale. At 100 million of network of network revenue, that's 4 million of payments. That's, you know, at the similar numbers, it's like three or four cents at three cents a token, which is dwarf the token price. So, if we can scale, this is pretty powerful upside. There's others that do lottery rewards. If you've lots of small users, you need to actually incentivize them more than you can pay per device. So, you need to lotteries are interesting way to do that. So, I think these are examples, Silencio, Huddle, Natix, where they have

thousands and thousands or even millions of phones, and you can't pay each of those phones enough to be really worthwhile. So, every month people are entered into lotteries to win, and that keeps people excited and interesting. So, suggest that this is the model I think works. You accept fiat, you um business customers stake rewards, you buy burn at a higher rate that's predictable. That's interesting. And you on the rewards you do you fiat link. And this is why I think DePIN wins, right? You've got services in critical markets like enabling AI. You've live networks, products that are working, that are cheaper and provide value. And you add that to powerful token economics, and you have the next generation of crypto. And I think that

is where we are with DePIN, and I think AI is going to and the DePIN together are going to accelerate and really be be the future. So, that's why I think DePIN is not a meme. And there were times when I think we kind of wished we were meme-like. I think now we're we're okay not being and I kind of miss some of the euphoria of the meme age, but I think important to realize this is real business that's working and doing incredibly well. So, take another look at the DePIN tokenomics report and look at DePIN DePINspace.co.

We've got a podcast with about 100 DePIN founders there if you're interested in the space. And we also do DePIN days. We've done I think 13 or 14 DePIN days with leading projects. But DePINspace.co is where you'll see that. And here's the DePIN tokenomics report. So, thank you very much for your time.