Part 1 of our thesis focused on supply-side characteristics of decentralized physical infrastructure networks. If you haven’t read it yet, check it out here.
A core premise of decentralized physical infrastructure networks (DePINs) is that offering network ownership (via token incentives) can help overcome coordination challenges that might otherwise hinder development of new infrastructure.
We’ve seen startups run this playbook effectively to build out the supply side of new infrastructure networks. But demand has historically remained more elusive. Luckily, there are signs that’s begun to change in meaningful ways over the past year. Pockets of real demand are emerging and beginning to provide a guidebook for markets that are well-suited to DePIN.
Here’s how we’re thinking about what’s most interesting on the demand side of DePIN projects.
Incumbent vs. Emergent Markets
Potential demand can be divided into two categories: incumbent vs. emergent markets.
Incumbent markets are ones where the demand for a service is well-known but is likely to be provided by existing businesses. Examples include telecom and wireless networks. Demand for such services is observable and quantifiable – the status quo shows that customers are already paying for these products. Consequently, it’s easier to size the opportunity. Typically DePIN projects in these markets aim to compete by offering lower-cost solutions with (sometimes) greater scale.
Emergent markets tend to be much less defined. Demand is typically created by some new technology or industry. Given that there is little to no existing infrastructure available to service this demand, DePIN projects are able to compete on a much wider array of vectors – including offering strong verification guarantees, identifying novel ways to build network capacity where centralized entities may be barred, and more (discussed in greater depth in part 1 of our DePIN thesis).
Emergent categories are far more interesting. There are more vectors on which DePIN can offer competitive, differentiated services.
Finding Emergent Markets
The next logical question is: where can we find emergent markets?
One useful proxy is to look to where startups are forming. Startups often emerge at inflection points: academic research becomes commercially viable and/or achieves a technological breakthrough, regulation changes, or we reach a tipping point in terms of a foundational technology’s distribution (e.g. a new piece of hardware). In recent years, we’ve seen such factors catalyze startup activity in industries like AI, energy, climate, genetics / bio, crypto, and spatial awareness (robotics, AR). And in general, new industries tend to be more experimental and accepting of alternative paths for capital formation, access to in-demand resources, and infrastructure development – opportunities for DePIN projects to potentially play a role.
At Variant, we’ve been especially excited about the potential for DePIN within AI and energy markets. Both categories are undergoing major technological transformations, with open questions as to the most effective structures for creating and distributing network outputs. In the case of energy, these questions involve whether to utilize local, distributed energy resources as a means for more efficient and stable power generation. Within the realm of AI, questions include which elements of the stack to open-source and whether decentralized infrastructure networks can produce performant, differentiated resources.
Both energy and AI networks also produce virtual resources. As a result, it’s much easier to tokenize and financialize the outputs. The stronger the tie between the network’s productivity and the token, the better the overall DePIN flywheel works.
Other Characteristics of Demand
Beyond emergent markets, there are a number of other qualities we look for on the demand side of DePIN projects.
Top of the list is the ability for stakeholders to participate as both consumers (demand) and infrastructure providers (supply). It’s well-demonstrated that DePIN projects can effectively use incentives to attract and coordinate supply. Converting that supply into active consumers – such as can be the case with distributed compute, model training, and energy networks – creates an initial base of adopters that helps battle test the network.
One area a little less exciting is data marketplaces. Crypto is very good at incentivizing quantity, but far less good at driving quality. Where DePIN within the realm of data does make more sense is when the consumer prioritizes a high refresh rate of new, granular data (e.g. commodities traders utilizing weather data).
All that being said, if the supply side offers a differentiated and better product than that of a centralized entity, it may still successfully attract demand regardless of whether the consumer traits map perfectly to our framework.
The holy grail of a DePIN is to offer a differentiated product, in an area where crypto uniquely solves a coordination challenge, with a category that has emergent consumers and product needs. If this sounds like what you’re building, reach out! I would love to chat.
Thank you to Jesse Walden, Dan Barabander, and Jack Gorman for feedback and discussions on this topic.