If you had pitched a “crowdsourced infrastructure” (later DePIN) to a room full of VCs in 2021, the conversation would have inevitably drifted toward token prices, staking APYs, and how fast you could bootstrap a map full of dots.
Fast forward to 2025, and the conversations I’m having (and the ones you should be having) are radically different. The DePIN sector has matured from a speculative $226 million niche into a $50 billion institutional giant. DePIN isn’t just building crypto-for-crypto projects. We are building the physical backbone of the cloud, AI and data economies.
But as the industry has grown, the startup playbook has changed as well. If you are still judging DePIN protocols by the same metrics you use for DeFi or memecoins, you are flying blind. Here is the reality of fundraising and building in the DePIN trenches over the last two years, and why the “growth-at-all-costs” era is dead.
From tech for pirates and public goods to critical infrastructure
The narrative shift has drastically changed. Between 2020 and 2025, we watched DePIN evolve from a proof-of-concept into a mature sector, even doubtfully projected to hit $3.5 trillion by 2028 (needs ~10x every year from now).
What changed? We are seeing over 1,500 active projects now. But the most exciting part isn’t the number of startups. It’s who is using them. We have moved past crypto-natives selling to crypto-natives. We have major enterprises like AT&T, ByteDance, and Cloudflare offloading data onto decentralized rails. This is the “institutionalization” phase we always dreamed about in crypto, and it is finally here. And it was reached way faster by DePIN than any other trend, but how do we really measure success today?
Why DeFi metrics are useless here
In DeFi, Total Value Locked (TVL) is a demi-god. If you lock in enough capital, you will eventually have a product. In DePIN, TVL is a vanity metric that often masks a zombie network.
I’ve seen projects boast about 500k+ nodes, but when you dig into the on-chain data or even just try to, the network is generating less revenue than a single garage sale. In the pre-2022 era, raw device counts were the primary signal of success. Today, a high device count without utilization is actually an expense. It means you are massively overpaying in token emissions for supply that no one wants.
DePIN is a hardware and logistics business wrapped in tokenomics. It requires a different lens. You can’t just fork a smart contract and liquidity mine your way to success. You have to secure customers and partnerships, and in the cloud business, deals typically take at least 6 months.
DePIN metrics that actually matter
If you want to spot the next unicorn, validate an existing project, or you are actually building one, you need to ignore the noise and focus on “industry-value-creation-first” projects. Here are the four realities that defined DePIN startup success in 2025.
The Holy Grail: Product-Market Fit or paid utilization
The single biggest red flag in DePIN is the abundance of supply. We have billions of devices online worldwide, but in many cases, no one needs just a device. One has to find a way to cleverly utilize them and ideally build a convenient service or API on top to make them valuable. The revenue doesn’t match the hardware deployment anymore.
Look at Filecoin. In 2022 (and I was working there as a Startup Operator), the network utilization was hovering around 3-4%. Critics called it dead space. But they focused on verifiable deals, and by Q3 2025, that number reached 36%, which is impressive for a project with 3EiB+ of storage available (3EiB = ~140 million text-only copies of the English Wikipedia).
As a founder, I can tell you: it is better to have 1,000 nodes at 80% utilization than 1,000,000 nodes sitting idle. Investors have wised up. They are looking for the revenue generation pathways for nodes connected. If your operators only make money from inflation, you are on a ticking clock.
The reality of price games vs enterprise reliability
Early DePIN pitches (mine included) often focused on price: “We are 90% cheaper than AWS!” And while that can be true for raw, interruptible resources (like the 95% savings on raw GPU hours), the enterprise reality is more nuanced.
When you start selling to serious companies, you are competing on reliability, not just price. An enterprise CIO will not risk their data to save a few pennies. They need SLAs.
The winning protocols, like Render, have to be ready for enterprise-grade reliability and SLAs. They aren’t trying to be free. They offer a sustainable 30-40% TCO (Total Cost of Ownership) advantage. This discount is massive enough to move the needle for a CFO but high enough to ensure node operators are paid fairly without infinite token printing.
Wallets don’t pay bills, enterprises do
Stop counting wallet addresses. It’s a noise metric. In the consumer crypto world, 100k active wallets is a triumph. In DePIN, one contract with a company like Lockheed Martin or TikTok is worth more than 50,000 retail speculators.
Chainlink is the prime example here. They didn’t just chase retail hype; they integrated with SWIFT and secured $95 billion in value. That is the moat. If your demand side is just other crypto degens, you are circular. If your demand side includes the Fortune 500, you are infrastructure.
Demand-first growth
Crypto infrastructure projects spent 2017-2022 solving the supply side. We proved we could use tokens to incentivize people to crowdsource hardware. Helium proved this with 342,000 IoT hotspots. The problem? We built bridges to nowhere.
Today, most of the ecosystem has already understood that supply is a solved game. Build a good product with good utilization and payouts, and supply will come. The startups winning now are building demand-first. Helium’s pivot to Mobile is the perfect case study. By focusing on where the paid data demand actually was, they drove annualized revenue to $18.3 million.
A humble 2025 DePIN startup checklist
If I were to align the DePIN industry on KPIs that matter, I would support startups focusing on demand metrics to see the project reality:
- Protocol Revenue (Organic): Show me revenue from fees, not token burns or staking requirements. Hitting $100k+ in monthly revenue is a stronger signal than $10M in TVL. (Titan Network has passed this point already.)
- Paid Utilization Rate: What percentage of your network is doing actual work? If it’s under 10%, you’re still a science experiment.
- Burn-to-Mint Equilibrium: Does the token model capture value? When a customer pays $1, does it reduce supply? Render and Helium have mastered this.
- Enterprise Value Add: Can you prove you are at least 30% cheaper than the centralized incumbent after accounting for reliability?
The long game: why we are just getting started
The industry has grown up. We have survived the bear markets and the “supply glut.” We now have regulatory clarity. The SEC’s Double Zero letter was a massive win, confirming that tokens for verifiable work aren’t necessarily securities.
This clears the runway for real adoption. We are seeing a surge in M&A, with traditional companies buying into the stack. The integration of DePIN into the global economy isn’t a “maybe” anymore; it’s an inevitability.
We are building the rails for the next decade of compute, energy, and connectivity. But this time, we are building it with revenue, not just roadmap promises.