Iteration is the Only Option

San Francisco, CA — March 20th, 2026

Token governance is broken, they say. We say, let's iterate.

Earlier this week, Tally announced that they were stepping out of the governance business. Tally, has been earnestly building in this space and we and the entire crypto community owe them a debt of gratitude for their contributions. When events like this happen, it triggers a reflection on the meta, the cycle and our place in history.

It was never going to be easy

There's a popular take circulating right now that crypto governance is dead. DAO governance is a failed experiment. The v0.1 was rough and therefore the whole idea should be abandoned.

We get it. And we understand why thoughtful people, including teams who've poured years into this space are stepping back. The cold hard truth is that it’s hard to build a business in this space.

We've seen this pattern before with the Internet in 2001, with mobile apps in the early 2010s, with crypto itself in every bear market and boom cycle. A new coordination mechanism launches, stumbles through its first few iterations, and people start to wonder if the whole idea was a dead end. Let's be honest, early token governance was rough. Low voter turnout, plutocratic dynamics, governance attacks, proposal fatigue, and decision quality that would make a homeowners' association blush. We don't dispute any of that. We lived it. We built through it.

But writing off DAOs because these experiments were messy is like writing off aviation because the Wright Flyer wasn't a 787. The first decentralized exchange wasn't Uniswap, it was Counterparty, built on top of Bitcoin around 2014. It took ten minutes to post an order. Ten minutes to get filled. Ten minutes to cancel. You'd sit there watching blocks tick by, hoping your trade would land before the price moved. It was, by any modern standard, ridiculous. If you had looked at Counterparty in 2014 and said "decentralized exchanges don't work," you would have been right, for a few years and wrong about the arrow of progress. A decade later, onchain exchanges process billions in daily volume, settle in seconds, and are credibly on the doorstep of Wall Street. Counterparty was the v0.1 of what Uniswap, dYdX, Aerodrome and Hyperliquid became. The idea was right. The execution just needed time, iteration, and builders who refused to quit.

Crypto governance is at its Counterparty moment right now. Whatever we build, will work out, as long as we keep iterating together.

Look around

Before we bury token governance, let's take a clear-eyed look at the institutions we're failing to live up to. Legacy institutions — the ones with centuries of iteration, trillions of dollars in infrastructure, and entire professions dedicated to their maintenance — are not exactly crushing it right now. Regulatory capture is the norm. Public trust in government is at historic lows. Corporate governance produces quarterly-optimized decisions that externalize costs onto everyone else. Congress has a lower approval rating than root canals. We're not saying DAOs are better than legacy institutions today. We're saying that if this is the bar — if this is what mature, battle-tested, well-funded governance produces — then maybe the critics should recalibrate their expectations for what "working" actually means. The difference between token governance and legacy institutions is that we can actually iterate. Our governance systems are programmable. We can upgrade contracts, adjust parameters, introduce new voting mechanisms, and deploy entirely new coordination primitives — all without waiting for a constitutional convention or a regime change.

The people saying "DAOs don't work" are pattern-matching on a snapshot. They're looking at a system in its infancy and comparing it to systems that have had centuries to calcify into the dysfunction we see today. Token governance is three years into serious experimentation. The US Congress is 237 years in and somehow still can't pass a budget on time.

We have so much work to do

Here's the part we actually agree with the critics on: there is an enormous amount of work to do. Voter turnout needs to improve. Delegation models need to get smarter. Proposal quality needs better infrastructure. Security and simulation tooling needs to be ubiquitous. Information asymmetry between delegates and tokenholders needs to shrink. The entire UX of participation needs to get radically better.

We know this because we think about it every single day. It's why Agora exists. It's why we build open source governance infrastructure, why we invest in simulation and verification, why we've built delegation systems, private voting, and notification infrastructure — not because it's easy, but because the alternative is ceding collective coordination to the same legacy structures that are visibly failing.

We know this because we think about it every single day. It's why Agora exists. It's why we build open source governance infrastructure, why we invest in simulation and verification, why we've built delegation systems, why we're working on private voting, and our notification infrastructure — not because it's easy, but because the alternative is ceding collective coordination to the same legacy structures that are visibly failing.

We're just getting started

Some say the experiment is over. We don't think so.

Token governance is the first programmable coordination mechanism in human history. It's composable, forkable, auditable, and upgradeable. It can learn from its mistakes faster than any institution before it — if we build the tooling to make that possible.

That's what we're doing at Agora. Not because we think DAOs are perfect today, but because we believe they can become dramatically better, faster than anything that came before them. The v0.1 was rough. The v0.2 is better. And we're just getting started.

If you're building in this space, we're glad you're here. There's so much work to do, and it's better done together.