Why 90%+ of Crypto Projects Die, and the Power Law That Explains Everything

This isn’t a motivational piece. This is an objective look at what separates the projects that survive from the ones that become cautionary tales. The mental heuristic you take from this lesson should fundamentally change how you evaluate any protocol, ecosystem, or token. Whether you’re building one, marketing one, or deciding whether to stay at one.

Power laws exist in basically every industry, but crypto makes them brutally visible.

The number one market player in any given segment captures 60-80% of the market share. Sometimes more. The number two might scrape together 10-15%. Everyone else fights over crumbs. This isn’t opinion. It’s a pattern that repeats across every vertical you can name.

Aave dominates lending. Competitive players like Morpho and Euler can chip away at individual liquidity pools or specific features, but they’re not taking huge market share from what Aave does well. Uniswap dominates DEX trading. Ethereum dominates L1 developer ecosystems. It even applies to marketing agencies. Very few take most of the clients and charge most of the money.

The takeaway is uncomfortable but necessary: if you’re not designing to be number one in your category, you probably shouldn’t even get started. Number 25 in a segment isn’t “room to grow.” It’s a slow death.

There are two growth trajectories in crypto. Once you learn to spot them, you can’t unsee them.

Healthy growth looks boring at first. Zoomed out, it’s a parabolic curve driven by network effects. Zoom in at any point and there are fluctuations. A dip here, a plateau there. But the overall direction is unmistakable. Each new user, each new builder, each new integration compounds the value of the whole system.

Ponzinomics looks exciting at first. Massive hype before launch. Botted Twitter accounts, fabricated testnet activity, random incentive programs, community “activations” that are really just noise. There’s enormous FOMO going into the launch, a spike, and then complete collapse. The fundamentals were never there.

Look at the L1 launches from early 2025 through now. You see this over and over. Projects hyped for three, four years. They launch. They collapse. The pattern is so consistent it’s almost funny, except real people lose real money.

From an ecosystem perspective (not consumer), there are three legitimate ways to capture value. Everything else is cope.

1. Transaction Fees

Ethereum charges gas. DEXs charge swap fees. PumpFun makes money. You provide a service, users pay for it, you generate revenue. Simple. Real.

2. Monetary Premium

Bitcoin, ETH, XMR. These assets have value in and of themselves because of what crypto was built to be: sovereign, undebaseable currency independent from jurisdictions and central banks. That philosophical value is itself a product. But most new crypto projects don’t have this, so don’t pretend yours does.

3. Governance Value

More contested, but tokens can carry governance value. Especially if the governance has rights to treasury, foundation creation, or meaningful oversight of the protocol’s future. When governance means something real, those tokens carry a premium.

Utility tokens that only exist to use one specific product. UseMyTokenToGoToMyCoffeeShop. Dead on arrival.

Governance tokens for dead protocols. Governing nothing is worth nothing.

Airdrop-farmed tokens that only incentivised pre-launch activity with no plan for after.

Most tokens don’t work. That’s why a growing number of crypto businesses are going the equity route instead. Projects with real revenue end up doing buybacks and burns to manufacture token holder value after the fact.

Slowly, slowly, then all at once.

Hayden starts researching in 2017. Gets a bit of angel capital. Launches at the end of 2018, right before DeFi Summer. It takes seven years before they’re hitting trillions in volume.

The joke is “zero marketing budget.” Obviously Uniswap has a marketing budget now. But in the foundation years? Nothing. They were pure innovation within an entirely new category, growing through mechanisms, not marketing spend.

Here’s what they got right and why it matters.

The AMM solved a real problem. It wasn’t a commodity, not a fork, not a copy-paste. Before Uniswap, we were using centralised exchanges and EtherDelta. The AMM was a genuine breakthrough.

They didn’t rush the token. They didn’t use a token to fundraise even though they easily could have run an ICO. They knew the token should come later, after the mechanism proved itself.

The network effect operates at the builder level. When protocols like Balancer, Compound, and dozens of others use Uniswap’s V1/V2/V3/V4 contracts under the hood, every LP in Uniswap’s pools gets used everywhere in DeFi. That’s where the moat gets built. Not at the consumer layer, but at the infrastructure layer. The users might not even know they’re using Uniswap. The apps built on top handle that relationship.

They built with authority and community. Working with early adopters, with Vitalik directly, designing the mechanism first, then going to market. An organic, value-first approach.

Compare this to SushiSwap, which forked Uniswap and tried to compete on limit orders. Uniswap had the deeper liquidity, the stronger network effect, and critically, the trust. Nobody’s using SushiSwap under the hood. It was and still is a purely consumer product. A fundamentally weaker position.

The same dynamic plays out with Ethereum and its endless parade of “ETH killers.” Since 2017, how many L1s have launched promising better speed, lower fees, flashier feature sets? It doesn’t matter, because Ethereum has the network effect. Developers trust it. The tooling exists. The liquidity is there. You can offer 20% yield like Luna did. I’d rather take the 4.5% and actually keep my money.

Liquidity and network effects. More liquidity means better prices, which attracts more users, which attracts more liquidity, which attracts more devs. This is the flywheel. This is what you want.

Developer ecosystem. More devs build more apps, which attract more users, which attract more devs. The devs are the founders. They build the applications that acquire users. The users end up using your infrastructure without you ever speaking to them or marketing to them directly. The user might not even know your infrastructure exists, because the apps are built on top.

Brand as Schelling point. “Just use Aave,” as Marc Zeller says. You want to become the default. You want to be Lindy. The thing that’s survived long enough that people assume it will continue to survive.

Regulatory capture. At scale, can you lobby? Can you influence policy in your favour? This matters for legitimate players.

Technical lock-in. If people are using you, is it cumbersome to switch? Do you make it really attractive to stay? This is a retention factor for ecosystem products.

Being faster. Being cheaper. Having better UX.

Going to Stanford. Having A16Z invest in you.

Claiming you’re “decentralised” with eight validators.

Having 90% mindshare on Kaito. Getting listed on Binance Alpha.

Having some niche feature nobody asked for. Like making backpacks for squirrels.

None of these are moats. That’s why so many projects that lean on these things still die.

Forget Discord users, Twitter impressions, and site visits. None of that actually matters. Here are the metrics that tell you if a project is succeeding or dying.

I don’t fully believe in the exact calculation, but the principle is vital.

Think of it like COVID transmission. Every infected person becomes a new node that can infect others, who become new nodes, who infect more people. That exponential spread is what you’re measuring.

In ecosystem terms: one partner leads to two, leads to five. It usually takes 20-30 partners before you see two or three standout winners building genuinely new primitives on your infrastructure. And these growth cycles take 6-18 months, sometimes 24, before the fog clears and you can see the real state of play.

Having $50M in TVL means nothing if nobody’s using it.

You always want to be measuring volume against TVL. The more capital efficient you are, the better your product. If liquid funds are dumping capital into your protocol but there’s no real usage, you’re not a good product. You’re a parking lot for mercenary capital.

Whether you’re an ecosystem, a protocol, or a DeFi application: if you’re not capital efficient, you’re not successful. Full stop.

This goes a step further. For every dollar of TVL, how much revenue are you generating?

If you have a million dollars locked and you’re making $10K on it, that’s a very different story from making $1M on it. You want the highest cash turnover possible on your locked capital. That’s the hallmark of any good protocol.

Yes, it matters too. Particularly how it’s designed. FDV, percentage of token float, emission schedules. Token price is important for marketing, for morale, for attracting talent. But it should be a consequence of the other metrics, not a substitute for them.

Critical for ecosystems.

If you’re Solana and you spend $10M a year through Superteam finding builders worldwide, it’s because a single anomalous 0.1% builder could produce billions in volume. The math works when you know what each new builder is worth.

If the builders coming in are generating revenue, getting users, and attracting funding, your ecosystem is healthy and your cost per acquisition is justified. If you’re spending $10-25K to acquire a good builder and they’re producing multiples of that in ecosystem value, you keep spending.

How much real capital is flowing in versus out? Net outflows since launch (like some protocols we’ve all seen) is obviously a bad signal.

How many builders are succeeding? Are VCs investing with conviction or just rotating capital? Are builders reinvesting revenue back into the ecosystem? What does the community actually look like beneath the surface?

Notice I haven’t mentioned impressions, follower counts, Discord members, or site traffic once. In theory, you could have 10 site visits, all 10 convert into builders, and the net marginal value is a million dollars per build. An incredible funnel doesn’t need big vanity numbers.

I don’t say this to be dramatic. I say it because the pattern is unmistakable after seeing it enough times.

If you need token price to go up to survive, you’re dead. A lot of teams are in this position. It’s basically impossible to recover from. Your product should work regardless of what the token is doing.

If you have no fees or revenue, you’re dead. Nobody is using you. You’re useless. Harsh but true.

If you’re not number one or two in your category, and don’t have a clear trajectory to get there, you’re dead. Cut your losses now. Number 25 isn’t a position you climb out of. Just stop.

If you need constant marketing just to stay alive, you’re dead. Marketing should be for scaling and growth after a validated product. If you need the marketing team at full blast just to keep the lights on, you’re as brittle as a drop-shipped white-label e-commerce product. One campaign goes wrong and you’re finished.

If VCs are your only users, you’re dead. This happens more than you’d think. One LP in a fund really likes the project, one venture partner gets excited, and suddenly the VCs become the main users. They share it among themselves, think it’s cool, but no actual users ever show up. VCs blowing smoke up each other’s arses isn’t product-market fit.

Here’s the heuristic that ties everything together.

Will your product survive a -90% drawdown?

Chainlink has been through them. Ethereum has been through them. Bitcoin has been through them. 90% drawdowns happen constantly in this industry. You don’t hope it won’t happen. You expect it and build accordingly.

When the trading fees dry up, when the market goes cold, when liquidity vanishes... can you survive and continue?

If the answer isn’t a clear yes, you have a problem that no amount of marketing will fix.

Winner takes the most. Design to be number one or pivot immediately.

Network effects are everything. If you’re not designing for compounding value through network effects, stop now.

Revenue always matters more than token price. If you’re not generating fees, utility, and capital velocity, the token catches up to that reality eventually. There’s only so much marketing can do.

Mechanism always beats distribution. You can get a billion impressions on TikTok. It will never replace a simple, elegant mechanism that compounds value over time and gives you that parabolic growth curve.

Survive the -90%. The ultimate test. Expect it. Build for it.

1. Map your protocol’s position on the Power Law curve. Go to Token Terminal or DefiLlama. See where your project (or one you’re studying) sits relative to the competition. Understand why they are where they are. If you were the CEO, what would you do to get to number one or two?

2. Identify the true competitive advantage. Not the features. If you had to create the killer offer to drive adoption, what would that actually be? Not “faster” or “cheaper.” The real thing.

3. Calculate your LTV and CAC using assumptions. If your CEO gives you $1M to go to market, how many builders can you acquire? What’s the expected value of those builders to the ecosystem? Think as quantitatively as possible.

4. Calculate ecosystem efficiency. Pick something like Aave. Go understand why they’re number one. Look at their history. Stani started with ETHLend, he was a lawyer, not a developer. How did a random lawyer from a Nordic country build the biggest decentralised lending protocol in the world? Why are they still number one despite Morpho and Euler? Is it just first-mover advantage, or is there more to it?