[Constitutional] AIP: Remove Cost Cap on Arbitrum Nova

1. Why I’m Voting FOR

  • We’ve covered about 284 ETH in Nova subsidies so far (86 ETH back in fees vs. 401 ETH in L1 costs – see chart¹).
  • Removing the cap means sequencer fees finally match real gas costs, so we’re not hiding the true economics.
  • That capital can be put to better use—like targeted grants, dev programs, or new scaling experiments.

¹ Hex Dashboard (“Cumulative Net” curve shows a steeper drop after EIP-4844 in March 2024)


2. Marketing & Strategy Perspective

Ultimately these subsidies were a growth campaign likely created to incentivize teams to come to Arbitrum Nova and spur more on-chain transactions (I assume there was a more specific goal).

I’d love to see how this campaign actually performed and what learnings we can apply to potential subsidies on Orbit. I think there’s a lesson here: treat Nova incentives like any growth campaign—set clear spend → acquire → retain metrics and iterate fast.

Key points:

  1. Builder acquisition: Which incentives (fee discounts, grants, hackathons) actually drove teams to Nova?
  2. ROI benchmarks: How much ETH did we spend per new dApp, per incremental TVL, or per transaction?
  3. Qualitative feedback: As we’ve discussed, I’d love to see direct interviews with early builders to learn why they chose Nova.

Questions for growth/ROI:

  • What KPIs (new builder count, active addresses, TVL lift) were we tracking, and what did they show?
  • Of those teams who joined, how many eventually moved to Arbitrum One or Orbit, and why?
  • Can we share these insights with the DAO to make Orbit-chain incentives more data-driven?