## Summary - Welcome credits: 2 → 1 (granted on first email verification) - Feedback credits: 2 → 1 (granted for submitting product feedback) - Total possible free credits per account: 4 → 2 ## Changes **Backend:** `NEW_USER_FREE_CREDITS` constant, feedback credit grant, AI evaluation prompt, model defaults **Frontend:** Landing page, dashboard banner, feedback dialog, auth dropdown menu text **Tests:** Updated all assertions and mock values (115 backend + 712 frontend tests pass) **Docs:** AI-CREDIT-GATING.md, API.md, README.md, PRODUCT-COMMUNICATION-GUIDE.md ## Test plan - [x] Backend tests pass (115 passed) - [x] Frontend Jest tests pass (712 passed) - [ ] CI pipeline passes - [ ] Verify landing page shows "1 Free Credit" after deploy - [ ] Verify new user signup grants 1 credit (not 2) @coderabbitai ignore 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
3.9 KiB
Product Communication Guide
Guidelines for how to communicate about Nomad Karaoke to users, in marketing materials, UI copy, and support interactions.
Core Principles
- Transparency over hype - Set realistic expectations so users aren't surprised
- Avoid AI terminology - "AI" has become a loaded term; say "our system" or "the tool" instead
- Acknowledge user effort - The human review step is required, not optional
- Highlight real differentiators - Real instrumentals (not covers), precise sync, multiple formats
Key Facts to Communicate Accurately
Timing
- Accurate: "Create a karaoke video in under 30 minutes"
- Avoid: "in minutes" (implies much faster than reality)
- Typical flow: ~20-30 minutes including human review
Song Availability
- Accurate: "Any song" - this IS true
- Users can: search by artist/title, paste YouTube URL, or upload their own audio
- If users can get the audio from somewhere, we can make karaoke from it
Lyrics Sync Quality
- Accurate: "Precise lyrics sync" - the timing/sync IS very accurate
- The sync (when words appear) is reliable and handled automatically
- The challenge is lyrics ACCURACY (correct words), not timing
Human Review Requirement
- Be clear: Users ALWAYS need to review the transcribed lyrics
- For songs with clear vocals (country, ballads): few or no corrections needed
- For complex songs: expect 5-10 minutes of corrections
- Frame as "Review & Correct" not "Preview and fine-tune" (the latter sounds optional)
Audio Separation
- Accurate: "Real instrumentals from the original song"
- Avoid: "Studio-quality separation" (oversells it)
- Key differentiator: NOT a cover band or MIDI recreation - it's the actual song
- Separation quality is very good but not perfect for all songs
Backing Vocals Option
- Users can choose between:
- Clean instrumental (all vocals removed)
- Instrumental with backing vocals preserved
- This matters for songs with backing vocals; irrelevant for songs without
Output Formats
- 4K MP4 karaoke video (main output)
- "With Vocals" version for sing-along practice
- CDG+MP3 ZIP for older/commercial karaoke systems
- All videos published to Nomad Karaoke YouTube channel (not user's channel)
YouTube Integration
- Accurate: Videos are published to the Nomad Karaoke YouTube channel
- Wrong: "Upload to your YouTube channel" - we don't do this
- This is for easy sharing, not user channel integration
Free Credits
- All new accounts receive 1 free credit to try the service
- No payment required to get started
- Credits never expire
- Additional credits available for purchase ($5/credit, bulk discounts available)
Terminology Preferences
| Avoid | Use Instead |
|---|---|
| AI, AI-powered, AI-generated | "our system", "the tool", "automated" |
| Perfect lyrics | "precise sync", "accurate timing" |
| Studio-quality | "real instrumentals", "actual song" |
| In minutes | "in under 30 minutes" |
| Fine-tune (sounds optional) | "review and correct" (sounds required) |
| Upload to YouTube | "published to our YouTube channel" |
FAQ Guidance
When answering user questions:
Q: Is it fully automatic? A: The transcription and sync are automatic, but you'll need to review and correct any misheard words before the final video is generated. For clear songs, this might take just a minute. For complex songs, expect 5-10 minutes.
Q: What songs work? A: Any song you can get audio for. Search our database, paste a YouTube link, or upload your own audio file.
Q: What do I get? A: Multiple files - 4K karaoke video, a sing-along version with vocals, and CDG+MP3 for older systems. Videos are also published to our YouTube channel.
Q: How long does it take? A: About 20-30 minutes total, including your review time.
Version History
| Version | Date | Changes |
|---|---|---|
| 1.0 | 2026-01-07 | Initial guide created based on founder feedback |