The Key to Building Sustainable DePIN

Many of the smartest people in crypto seem to think that Decentralized Physical Infrastructure Networks (DePIN) will be among the first use cases to make crypto mainstream. They are predicting a tidal wave of growth, similar to NFTs or stablecoins before it.

That’s because DePIN can accomplish simply remarkable feats. By using crypto incentives to align millions of participants, DePIN projects can build net new products that simply weren’t possible before.

Ariel Seidman is the co-founder and CEO of Hivemapper, the fastest-growing mapping community in the world.

In December, Helium Mobile rolled out a $20-a-month unlimited cell plan nationwide powered by community-owned 5G hotspots. Render Network created a robust, permissionless ML cloud during the world’s GPU crisis in less than six months. And, my company, Hivemapper, mapped 10% of the world’s roads in less than a year. These things simply weren’t possible before DePIN crypto-economics.

DePIN is an exciting vision of the future, but for it to be more than a fad new projects must learn from the first movers in the space. The first wave of DePIN projects experimented with tons of different models, made lots of mistakes, and learned from them. Yet these learnings seem to have disappeared down the memory hole. And, absent that context, new DePINs risk creating a boom-and-bust cycle that would leave a black mark on the entire sector.

To help DePIN chart a more sustainable path, I’d like to share some of the most important lessons we’ve learned at Hivemapper along the way. More broadly, for anyone looking to participate in DePINs, these leanings are also critical points to understand when evaluating whether a DePIN project is likely to achieve long-term success.

The Static Rewards Trap

Tokens gamify rewards in such a way that encourages contributors to collectively build a product that’s greater than the sum of its parts. On the other side, usage of the product results in tokens being consumed.

To be sustainable, a DePIN project needs a sound strategy for generating supply, and, perhaps more importantly, genuine demand for the product. Without that, the demand for the token is purely speculative — a memecoin, not a useful infrastructure project that can impact the lives of billions.

With the right tools generating supply is the easy part: give tokens for work. The tricky part is figuring out how many to give. Not so many that you have none to give at later stages of growth, but also not so little that no one shows up to work. New projects also need to be especially mindful of the “cold start” problem: a new network can’t offer much value with one contributor, or even a hundred; it needs to reach some degree threshold-scale before even thinking about servicing demand.

Generating genuine demand is fundamentally harder — you have to find product-market fit, and fast. In our experience, it helps to focus on a very real pain point that a decentralized approach is uniquely positioned to fix. It’s tempting to take shortcuts, and the most common shortcut we’ve seen is adopting static rather than dynamic rewards. We call this the “Static Rewards Trap.”

It’s way easier to attract contributors in the beginning if you let them earn rewards just for engaging with the project, regardless of how much value they add. But static rewards are fatal in the long run because they break the network’s underlying incentive structure.

Imagine if Helium gave the same static rewards for every 5G hotspot whether it was installed in a place where people live and work, like Manhattan, or a sparsely populated area, like Death Valley. There would be no incentive to enhance coverage in the places where people need it most, hindering its mission of democratizing access to communication and beating centralized cell providers.

Imagine if Hivemapper issued the same static rewards for every dashcam. It would be a disaster. People could get rewards by putting dashcams in any car, even an extra one sitting in the driveway that barely gets driven. They could put five or 10 dashcams in the same car, raking in extra rewards while capturing the same map data over and over again.

If DePIN projects treat all contributions equally — giving out rewards everywhere — they will fail to achieve critical mass anywhere. Productivity needs to be a protocol’s mantra: establish baseline reward for the least useful contributions then obsess over quality and outright reject duplicative work or low-quality data.

Using a utility token to align incentives is the superpower of DePIN. But if a project falls into the Static Rewards Trap, contributors have little incentive to help make the product better long term. Don’t trade your enduring superpower for short-lived growth.

Exploring the Design of Dynamic Rewards

There are four main dimensions for dynamic rewards across most DePIN projects. They are:

Geography (Infrastructure is more valuable in some places than others)

Contributor Productivity (Higher-productivity contributors deserve more rewards)

Contributor Quality (Higher-quality contributions deserve more rewards)

Network Progress (A more useful network should generate more rewards overall)

When designing the Hivemapper Network, we didn’t get all of this right. The Hivemapper Foundation has been the first to admit it made errors, such as issuing excessive bounties to incentivize mapping in specific regions of the world. But we’ve made steady progress, helping us to cultivate one of the most productive communities in DePIN.

To put this in practice, here’s how we used this framework for our network:

Geography: We designed Hivemapper so customer demand in a region would translate into better rewards, with 100% of tokens burned for map usage being re-minted to the contributors who contributed the data. We’ve made similar changes to minted rewards as well, adding a rewards multiplier for regions of the world that collect more useful map data – even before customer usage ramps up. In their brilliant essay on the design of DePIN networks, our friends at Multicoin Capital called this the “land and expand” approach.

Contributor Productivity: As a Hivemapper contributor, rewards are mostly determined by how far you drive, and the uniqueness (or “freshness”) of the roads you drive. It would be much easier to onboard contributors if we didn’t do it this way. Some people don’t drive much. Some people mostly drive the same roads while commuting to work. These people won’t generate as much rewards from Hivemapper, and that’s how it should be. For this reason, Hivemapper has been focused on recruiting gig drivers, truckers and commercial fleets. These drivers are highly productive, passively collecting map data as they go about their job – and reaping the best rewards.

Contributor Quality: Building a map requires high-quality data with precise positioning. For this reason, Hivemapper has implemented reputation scores for each type of contributor. The higher the contributor’s reputation, the better the rewards they receive. Done well, this correctly aligns incentives. Hivemapper contributors are eager to mount their dashcams correctly for the best image quality and the best GPS signal for accurately mapping locations. They are serving the needs of the network in order to maximize their own rewards.

Network Progress: The size of Hivemapper’s weekly rewards pool is based on a dynamic metric called Map Progress, which measures how much mapping has been done across the network. Many projects use a fixed minting schedule, which results in heavily front-loaded rewards for early adopters. It’s true that early adopters deserve bigger rewards for taking greater risk and dealing with more technical issues. However, a fixed-minting schedule results in ever-falling rewards as more contributors come online – just when the network is getting more useful! It can also result in toxic relations between early adopters and later contributors, creating an unhealthy distraction.

We’re open-sourcing this framework because it helped us build one of the fastest growing communities in DePIN. We’ve been able to do more with less, mostly because we’re standing on the shoulders of giants before us, and we hope to repay the favor.

Incentive design isn’t easy. But if DePIN builders remain focused, and if stakeholders hold them accountable, we can live up to the potential of DePIN and build a world where more of humanity’s critical infrastructure is owned and operated by the people that use it.

Source