Five Months Operating the Cactus Platform

In May we shared what six weeks of operating Tally's governance platform looked like. Since then, Tally has rebranded to Cactus while we remained committed to stability and continuity. If you use Cactus often, you should have noticed it's faster and steadier than it was, without any disruptive changes. Which is how we wanted this new beginning to be.

Steady doesn't mean idle, though. Since the rebrand we've shipped a large batch of performance, correctness, and reliability work, along with some small features and quality of life improvements. Let's dig into that and a few other updates.

What we've shipped

Everything should be easily findable.

Global search is now a click or ⌘-K away. Type anything to search DAOs, proposals, and delegates; navigate with your arrow keys and smash enter to navigate.

Beyond global search, we've improved searchability of data across the app:

• Search proposals and filter by status and type, with infinite scrolling through even the largest governance histories.
• Delegate profiles now show each delegate's share of voting power, so you can see who actually moves outcomes.
• Cactus remembers recently visited DAOs and gives better guidance when a search comes up empty.
• Treasury sources are searchable by name or address, with duplicates clearly identified.

Need for speed.

• Pages open quickly: DAO discovery loads instantly, the homepage and proposal pages ship less code and do less work upfront, and heavier charts load only when you ask for them.
• Data arrives sooner: for supported EVM governance, Cactus now indexes directly from the chain head on a 10-second cadence, rather than the roughly three-minute delay we ran before. That delay was our buffer against chain reorganizations, but we now handle reorgs directly so the buffer is gone.
• Delegate search is over 10x faster in the cases we measured, and even very large delegation histories load quickly.

We made data more trustworthy.

• Quorum is now read directly from governor contracts, missing proposal titles were repaired, and vote, delegate, and token-holder totals are more resilient.
• When we don't know something, we say so: treasury views show honest refresh and error states instead of silently displaying zeros when Safe data is unavailable.

And we made acting on governance smoother and safer.

• For DAOs with gasless voting and delegation, it stayed free and simple while we strengthened the signing infrastructure underneath.
• Delegation now continues cleanly after a network switch.
• Proposal drafts edit faster and preserve your changes. Queued proposals show clear indexing states and cannot be executed until their timelock is ready.
• Proposal pages are calmer and more stable, especially on mobile, with fewer loading flashes, layout shifts, and stale states, so you're never left wondering whether your action registered.

We also shipped plenty of invisible reliability work, namely stronger database monitoring, a cleanup of unused networks, and the retirement of archived legacy infrastructure. All while preserving historical governance records for active DAOs.

What we've been hearing

We've continued in direct conversation with many DAOs, delegates, legal providers, custodians and researchers since the last update. A few new patterns stand out:

Legal considerations can make it or break it. The Clarity act is all but dead, but a bevy of proposed rulemaking from regulatory agencies is opening up new possibilities. For now, DAO launches have slowed: everyone has been waiting to see what the rules would be. While we probably won't get Clarity with a capital "C", we may soon get enough clarity to get teams off the sidelines. Regulatory arbitrage may have motivated some DAOs in past cycles, but regulatory certainty may be what drives the next wave.

Continuity assurances are key. With at least a couple of governance platforms having been wound down or deprecated this year, many DAOs were forced to migrate on short notice and didn't like the experience. DAO operators are chatting about redundant frontends but that can also be expensive or inefficient. We're studying ways to address this beyond our own commitment to Cactus in the long term, and this matter is shaping how we run the platform. This is key because delegates, for their part, vote where habit lives and aren't enjoying the different experiences of custom frontends.

Governance security remains bonked. Conversations are either focused on the need for monitoring or the need to use updated governance systems, e.g. proper security councils, proposer roles, good old timelocks and optimistic governance solutions. We're exploring all of these, and we want Cactus to foster adoption of better security. There's also interest in shielded voting onchain, and we're exploring that too.

There's a lot to explore governance-wise. The most interesting requests we're getting aren't about proposals or standard platform features. But about how onchain organizations can better coordinate at scale over the next decade and beyond. Cactus emerged as a brand deliberately focused on onchain organizations, and not just on being a better voting platform. We're looking forward to having more of these conversations.

What's next?

We'll have more news about the domain migration process soon. For now the app still lives at the old tally.xyz domain, and we'll start forwarding to cactushq.xyz in the coming months. Old links will keep working for a while.

Beyond that, you can expect continued steady improvements to performance and usability, combined with a host of new features, big and small, shipped at an accelerating cadence. Behind it all, we have a singular focus: help stakeholders grow dynamic onchain organizations without sacrificing on transparency, security, and capture resistance. Everything is in service of that.

Lastly, as we've said before, we want the broader DAO community informing where this goes. If you'd like to be part of that conversation, reach out. There's lots left to do, so for us: back to work!