Ordered by engagement-winning value, not standalone app revenue, because Twenty has no developer payment rails — all six monetize as services.
- 06 Migration (HubSpot only) — only idea validated by money changing hands today. Best unit economics; least SDK-coupled. Do not build Pipedrive/Salesforce speculatively.
- 01 Lead scoring — cheap (2-4d), zero external deps. Value is as a sales demo.
- 03 PostHog — upsell inside every #1 engagement (its buyer is a HubSpot refugee).
- 02 WhatsApp — highest ceiling, highest risk. Start Meta Business verification paperwork in parallel with #1 (zero eng cost); reassess position when it clears.
- 05 E-signature — PandaDoc first (API key, no JWT).
- 04 Google Contacts/Tasks — consider not building. Lowest urgency in the set.
Decision: publish apps free/open (MIT) as distribution + credibility; sell implementation and retainers. Do not build a paid feature tier initially.
Why:
- No platform payment rails; no revenue share; no pricing field in the manifest.
- Freemium needs 10k+ free users / 18-36 months. Twenty's whole ecosystem is 18 apps.
- AGPL SDK bundles into the shipped artifact with no linking exception — a closed-source paid tier is legally unresolved, and the npm/marketplace path publishes source anyway.
- The two highest-value PRDs (06, 02) are not freemium-shaped: 06 is run-once per engagement, 02 is gated on client-side Meta verification.
If a paid tier is added later: make it a hosted service the app calls, not a license check in shipped source. Sidesteps both the AGPL question and the fork risk.
Deferred, not decided: anti-piracy/DRM posture. Research on enforcement effectiveness was retracted as unreliable — no sourced basis either way. Moot while there is no paid tier; revisit only if one is added.