Executive Summary
Arbitrum Unlocked 2.0, led by Castle Labs and funded under the Arbitrum Education, Community Growth and Events 3.0 domain, has successfully delivered a six-month educational campaign focused on Arbitrum’s ecosystem, tech stack, and DAO governance. The campaign applied the core learnings from our first Arbitrum Unlocked grant, including a bi-weekly cadence, an ecosystem-weighted content mix, and the removal of X Spaces in favour of data-led research.
Over the course of the campaign, we met 100% of our milestone deliverables:
- 12 Newsletter Segments (Arbitrum Corner), with 9 corresponding X Articles
- 6 Feature Articles
- 3 Data-Led Research Reports
In aggregate, Arbitrum Unlocked 2.0 generated:
- 487,453 total X impressions (108% of the 450k target)
- 33,949 unique newsletter opens
- 842 Substack clicks
- 54 total content pieces across the campaign, split between 27 committed milestone deliverables and 27 bonus pieces published responsively alongside the roadmap
Alongside the campaign numbers, Castle’s audience grew meaningfully, albeit through several activities outside of this grant:
- Substack subscribers grew from 3,700 to 6,700
- X followers grew from ~10,800 to ~14,600
The main takeaways from this campaign are:
- Data-led Research Reports were the strongest format, averaging 23,052 X impressions per piece and carrying the highest engagement rates
- Ecosystem content once again led on reach at 57% of total impressions, confirming the 1.0 finding that ecosystem-level coverage is the highest-leverage category for Arbitrum mindshare
- Bonus, non-milestone content delivered 46% of total campaign impressions across 27 additional pieces, roughly as much as all six milestones combined
- Through a noisy Kaito-driven year on CT, Castle’s Arbitrum coverage matured into one of the ecosystem’s most referenced editorial voices, with our content rivalling official channels relative to our size
Overview and Milestones
Arbitrum Unlocked 2.0 delivered 54 pieces of content across six milestones and five formats. Every Feature Article and Research Report was repurposed as an X Article, with Tweets and single posts extending distribution into the X timeline.
|Format|Count|
|Weekly Newsletter Segments|12|
|Feature Articles (milestone and bonus)|12|
|Research Reports (milestone and bonus)|5|
|X Articles|10|
|Tweets|15|
|Total|54|
The campaign was structured around six monthly milestones. Milestones 1, 2, 3, and 5 were completed on schedule between August and December 2025. Milestones 4 and 6 extended into January to March 2026, a scheduling shift driven by the extension of DRIP Season 1 (which we were tracking for the retrospective report) and the winter holiday publication window. All deliverables were completed in full, and all headline performance targets were met.
Milestone 1
1 Feature Article, 2 Newsletter Segments
- Feature Article: Arbitrum’s Journey Bootstrapping Its RWA Ecosystem, 52,322 impressions
Milestone 2
1 Research Report, 2 Newsletter Segments
- Research Report: State of Stylus: Breaking Barriers on Arbitrum, 32,892 impressions
Milestone 3
1 Feature Article, 2 Newsletter Segments
- Feature Article: Drip or Drop? Analysing the Effectiveness of the DRIP Incentive Program, 27,289 impressions
Milestone 4
1 Research Report, 2 Newsletter Segments
- Research Report: Arbitrum Today: Strategic Focus and Growth Drivers, 22,478 impressions
Milestone 5
1 Feature Article, 2 Newsletter Segments
- Feature Article: RedStone x Arbitrum Stylus: Oracles Beyond the EVM Limitations, 25,496 impressions
Milestone 6
1 Research Report, 2 Newsletter Segments
- Research Report: Inside Arbitrum DRIP Season 1, 4,989 impressions
Data, Metrics and KPIs
The primary metrics tracked for this grant were:
- Number of deliverables successfully submitted
- X Impressions
- X Engagement
- X URL Clicks
- Substack views, opens, and clicks
The full dataset, including all publication-level performance data and chart visualisations, is available at: docs.google.com/spreadsheets/d/1Pakrod0A8qAfNLgSaUUJisLo1QofXf4NwxUkeEdtI6Q
Campaign totals against targets
| Metric | Target | Actual | % of target |
|---|---|---|---|
| X Impressions | 450,000 | 487,453 | 108% |
| Substack Opens (total) | 22,500 | 33,949 | 151% |
A note on the Substack target. Our original grant application set a 45,000 opens target by extrapolating the 1.0 campaign’s figures over a six-month window, but that number implicitly carried forward a weekly cadence. The 2.0 campaign ran on a bi-weekly cadence by design (a key 1.0 learning), which means the comparable target should have been roughly half: 22,500. Against that target, the 33,949 opens figure represents a 151% delivery.
Performance by Format
Research Reports were the standout format. Averaging 23,052 X impressions per piece, they delivered the highest engagement rates and click-through rates of any format in the campaign. This was the central 2.0 hypothesis in action: replacing AMAs with data-led research produces longer-shelf-life content that earns amplification on its merits rather than on live-audience novelty. The strongest individual Research Report was the State of Stylus analysis at 32,892 impressions.
Feature Articles were the highest-volume long-form format, averaging 19,209 impressions across 12 pieces (milestone and bonus combined). The three milestone Feature Articles averaged 35,036 impressions, with the RWA deep-dive leading at 52,322 impressions.
Newsletter segments provided consistent touchpoints with our Substack subscriber base, averaging 2,440 X impressions and 844 Substack opens per segment. X Articles acted as the primary distribution layer for long-form content, averaging 7,029 impressions per piece. Tweets and single posts filled the timing gaps between longer deliverables, averaging 2,808 impressions.
Performance by Tag
Ecosystem-tagged content led overall reach, generating 279,032 impressions across 36 pieces, or 57% of total campaign impressions. This confirmed the headline finding from Arbitrum Unlocked 1.0: ecosystem coverage, particularly protocol spotlights and sector deep-dives, is the highest-leverage category for sustained mindshare on Arbitrum. The campaign’s highest-performing single piece came from this category, a comparative ecosystem analysis published early in the campaign: https://x.com/castle_labs/status/1950178494761091373
DAO-tagged content performed strongly in its own right, with 12 pieces generating 134,011 impressions (27% of total). DRIP analysis and incentive programme research drove most of this reach, which speaks to sustained reader interest in governance outcomes when the analysis is data-backed and independent of the DAO itself.
Tech Stack content contributed 74,410 impressions across 6 pieces. Lower on volume than Ecosystem or DAO, but the Tech Stack category consistently reached developers and protocol teams who are difficult to capture through other formats. The Stylus coverage in particular, both the State of Stylus report and the Stylus Demo Day X Article, was where this showed up most clearly.
Bonus Content
Beyond the 27 committed milestone deliverables, Castle produced an additional 27 bonus pieces across the campaign, generating 225,981 impressions, or 46% of total campaign reach. That is roughly the same total impression volume as the six milestones combined. These pieces were not tied to specific milestones and represented no additional cost to the grant. They reflect Castle’s broader, day-to-day coverage commitment to Arbitrum, and our ability to respond to protocol launches, governance moments, and tech releases as they happen rather than when the next milestone is due.
The bonus content clustered across four strategic themes:
Protocol launch and growth support
Timely coverage for emerging and growing protocols, timed to major announcements. Highest performers included a comparative ecosystem analysis at 51,373 impressions (the campaign’s top single piece overall), Boros: The Final Piece of Pendle’s Yield Trading Ecosystem at 24,258, Fluid: DeFi’s New Dominator? at 19,725, USDai: Fuel for the Digital Age at 20,557, and Variational: Secrecy and Grand Designs at 17,655.
DAO governance and incentive programme analysis
Original research and real-time analysis of DAO initiatives. DRIP & Incentives Outcomes (10,459 impressions), The Builder Momentum: How Chain Adoption Correlates with Small Builder Activity (9,944), and User Retention Analysis for Incentive Programs (8,808) gave readers early, structural analysis of DRIP’s impact and actionable signal for future DAO programmes.
Tech stack and infrastructure
Coverage of scaling and developer tooling, including Arbitrum Orbit, StylusPort, and the Arbitrum Audit Program.
Gaming and ecosystem milestones
Coverage of Arbitrum’s gaming vertical including the Gaming Ecosystem Overview and Proof of Play: Building “Forever Games”, alongside real-time coverage of the $4B TVL and $1B in RWAs milestones, and the Ronin’s Return to Ethereum research report.
Bonus content like this only works when the coverage team is already in the market every day. Milestone-funded work gave the campaign structure. Everything else came from the fact that Castle was going to be writing about Arbitrum regardless.
The bonus content programme is the clearest illustration of what it means to have a dedicated ecosystem partner embedded in the market. Castle was able to publish responsively because our team was already watching the ecosystem every day for our core editorial operation. Grant-funded deliverables gave us structure; proximity and editorial infrastructure gave us the capacity to go above and beyond where it mattered.
Learnings
Data-led research travels further than any other format
The 2.0 hypothesis, that swapping AMAs for data-led Research Reports would produce more durable value, was validated. Research Reports averaged 23,052 impressions per piece with the highest engagement and click rates of any format in the campaign. Dune dashboards and original quantitative work proved reusable in ways we did not fully anticipate: data visualisations circulated independently on X, extending their reach well beyond the reports’ publication week.
Ecosystem remains the highest-leverage tag
The 1.0 finding that Ecosystem content outperforms held again through 2.0. Ecosystem-tagged pieces generated 57% of total impressions across roughly two-thirds of content output, and consistently attracted the strongest amplification from protocol accounts and the community.
Distribution timing matters more than we’d like
The DRIP Season 1 retrospective is the clearest example. We published in close proximity to Entropy’s DRIP analysis, and the overlap meant our piece received less amplification from Arbitrum-affiliated accounts than it otherwise would have. Engagement per impression held up, but raw impressions were suppressed. For future campaigns, better coordination with the Foundation and other research providers on publication timing would reduce this kind of self-cancellation.
Responsive publishing compounds grant ROI
27 bonus pieces contributing 46% of total campaign reach are structural, not coincidental. Grants that fund only fixed deliverables from partners without a broader ecosystem presence will not see this compounding effect. A small flexibility budget layered on top of fixed milestones would meaningfully increase overall content ROI, provided the partner has the editorial infrastructure to use it. Castle is in that position by design.
Consistent editorial voices earn their keep in a Kaito cycle
The second half of 2025 saw Arbitrum’s mindshare become increasingly fragmented. Kaito leaderboard dynamics pulled incentives toward volume over substance, and while yappers boosted top-line mindshare numbers, a large share of the resulting content was superficial or duplicative. Price action got covered; the tech stack, the DAO, and the ecosystem’s structural developments got less attention than they deserved. Arbitrum’s real technical advantages (Stylus, BoLD, Orbit chains, the depth of its DeFi liquidity) risked being crowded out by noise.
Interestingly, individual contributors from Castle Labs were consistently among the top 10 on the Kaito Arbitrum leaderboard throughout this period, even though organisation accounts like @castle_labs do not count toward that leaderboard by design. Throughout the cycle, Castle operated as a stable editorial voice covering all three pillars of Arbitrum with original analysis rather than recycled takes. A number of our pieces were referenced by other creators and Arbitrum Ambassadors as source material, which is the best signal we could ask for that substantive coverage still has a role to play in a Kaito-saturated environment.
Acknowledgements
Castle’s work this campaign was consistently amplified by Arbitrum-aligned entities (AAEs), delegates, and community voices who shared, quote-tweeted, and pushed our content to new audiences. We appreciated it every time, so thank you.
Foundation Collaboration and What’s Next
Over the course of the campaign, Castle held several conversations with the Arbitrum Foundation about closer collaboration on ecosystem communications. Those conversations didn’t convert into a formal engagement, with the Foundation choosing to keep more of that work in-house. We remain available and happy to support Foundation-level or other AAE initiatives should priorities shift.
With Arbitrum Unlocked 2.0 now wrapped, and no further grant application planned on our side, Castle’s Arbitrum coverage will move back into the rhythm of our regular editorial calendar. We’ll continue to cover Arbitrum when the ecosystem does something our analysts find interesting, and we expect that to happen often. But the volume of dedicated coverage during 2.0 reflected a grant-funded commitment, and readers should expect the overall share of Arbitrum content in the Chronicle and on our X channels to normalise accordingly. We want to be straightforward about that. Arbitrum is a core ecosystem for us and will continue to feature, just at a cadence set by editorial interest rather than a fixed grant roadmap.
Conclusion
Castle Labs has completed the Arbitrum Unlocked 2.0 educational campaign in full, against the milestones defined in our application and the broader goals of the Arbitrum Education, Community Growth and Events 3.0 domain.
The campaign centred on high-quality, data-led educational content across the three pillars of the Arbitrum ecosystem: highlighting protocol developments and sector growth, abstracting the tech stack for a broader audience, and providing independent analysis of DAO governance and incentive programmes. The 2.0 structure of 12 newsletter segments, 3 feature articles, 3 data-led research reports, plus 27 bonus pieces, produced 487,453 impressions, exceeded the headline KPI target by 8%, and delivered a 100% milestone completion rate. Alongside it, our Substack base grew from 3,700 to 6,700 subscribers and our X following reached around 14,600.
More broadly, the campaign confirmed Castle’s role as one of the most consistent, substantive editorial voices covering Arbitrum, a role that grew in importance as the broader mindshare environment became noisier throughout 2025. Our thanks to the Arbitrum Education, Community Growth and Events domain and the SEEDGov team for their support throughout the campaign, and to the Arbitrum community for reading, sharing, and engaging with the work.