Anybody dealt with a DAO where Discord is active but on-chain activity is flat?

We faced this exact issue in a 20K-member DAO. The solution wasn’t changing KPIs first — it was redefining intent tiers. We mapped contributors into four buckets:

1. Observers (wallets seen, zero interaction)

2. Light participants (read proposals, attend calls)

3. Active voters

4. Contributors (working groups, grants, execution)

Then we mapped every community action to the tier it actually influences.
Memes → Tier 1
Townhalls → Tier 2
Voting guides → Tier 3
Contributor onboarding → Tier 4

Suddenly the community team understood why “big Discord months” didn’t move governance metrics — their activities were strengthening Tier 1 only.

We then built KPIs around tier progression, not raw engagement volume.
This aligned both teams without blaming anyone for the earlier mismatch.

Try building an intent funnel first — KPIs make more sense once everyone agrees on contributor stages.

There’s a known pattern in DAO data: less than 15% of social engagement correlates with any on-chain behaviour (Messari’s contributor funnel study). This is why active Discord ≠ active governance.

What helped us was defining three shared North Star metrics that both community and growth teams influenced:

• Wallet retention (returning wallets month-over-month)

• Governance conversion funnel (proposal readers → commenters → voters)

• Working-group depth (contributors active for 3+ cycles)

Social metrics weren’t removed — they just became inputs, not outputs.
So the marketing team could still try creative experiments, but success was measured by whether those experiments nudged someone further down the governance funnel.

A practical tip:
Plot Discord engagement next to wallet retention.
If the lines diverge, the misalignment becomes visually obvious, and conversations become much easier.

Your team isn’t disagreeing on data — they’re disagreeing on what “community” means. Many DAOs still treat community as emotional energy and governance as rational behaviour. But in practice, both are measurable parts of the same contributor system.

We introduced a Governance Health Index (GHI), a single composite score shared by both teams. It included:

1. Participation consistency

2. Wallet diversity

3. Proposal review depth

4. Working-group retention

5. Cross-channel behaviour (Discord → Snapshot → on-chain)

When both community and governance were accountable for the same GHI score, debates stopped.
Marketing still focused on vibe-building, but they designed activities that encouraged movement toward governance touchpoints rather than isolated social engagement.

If your treasury renewal is coming, a shared framework like GHI will help you make the governance risk visible before it becomes political.

From a systems-thinking view, you’re optimising two subsystems with unrelated feedback loops. Social engagement reacts instantly; governance activity moves slowly. Unless you build one shared contributor funnel, noise will always rise faster than signal. Start by mapping which Discord behaviours actually precede Snapshot or on-chain actions. The dependency graph itself becomes the alignment tool.

We solved this by doing a “noise audit.” Every metric had to tie back to a wallet or governance action. Once we removed bots, airdrop farmers, and automated posts, over half the Discord activity vanished — and suddenly everyone saw why social KPIs were misleading. Honestly, cleaning the dataset did more to align our team than any meeting.