# 🎯 Channel Plan — Luna's Birthday

**Date:** 2026-08-27
**Owner:** Tommy Tornroos
**Inspired by:** EpicHappyBirthdays (1,000+ name videos, A-Z directory)
**Target launch:** Q4 2026 (after Tommy green-lights music path)

---

## 1. Niche Thesis

### Why "Happy Birthday [Name]" is a sleeper-blue-ocean niche

1. **Search intent is exact-match and emotional.** People search `"happy birthday sarah"` or `"happy birthday liam"` once a year for that person. The intent is "I need a video I can text to them RIGHT NOW." That's the highest-intent, lowest-decision-latency search on YouTube.

2. **Volume is enormous and atomic.** The US alone has ~3.6M babies born/year (CDC), and another ~330M living people with first names. The top 1,000 names cover ~70% of the population (US Census Surname/Name data). At 365 names/year, we cover the top ~5% of names in Year 1 and tap the biggest search clusters first.

3. **Production is templateable.** Every video is the same 30-sec structure with a name slot. Once we have the template + an asset pipeline, each new video is ~5-10 min of human work (name slot fill + render + upload).

4. **Lifetime cumulative traffic.** A birthday video for "Sarah" gets viewed every day of every year — by every Sarah-celebrant from now until the heat death of the internet. The aggregate impressions compound forever. A 2015 EpicHappyBirthdays Sarah video (the cat version) still racks up views in 2026 — search engine results don't decay the way news does.

5. **Gift-grade emotional hook.** People share these in family group chats. The free distribution (people forwarding the URL to aunties and grandparents) does our marketing for us. Every share = one new viewer who didn't have to search.

### EpicHappyBirthdays — current state (researched 2026-08-27)

| Signal | Value | Source |
|---|---|---|
| Channel handle | `@EpicHappyBirthdays` | youtube.com/@EpicHappyBirthdays |
| Catalog size | "Over 1,000 personalized and custom Happy Birthday Songs" | Channel banner / about text |
| Companion site | epichappybirthdaysongs.com | A-Z name directory + /ages/ index |
| Multi-variant per name | 4-5 versions per name (rock, CAT, SLOTH RAP, ALIEN, "this is your gift") | epichappybirthdaysongs.com/sarah/ listing |
| Companion monetization | Bandcamp (per-name albums), Spotify (artist profile), TikTok 17.5K followers | bandcamp.com, tiktok.com/@epichappybirthdays |
| Premium upsell | /custom-song-video/ page — bespoke commissions | epichappybirthdaysongs.com/custom-song-video/ |
| Sample videos (cited) | [Sarah](https://youtu.be/76qbPWmrr8o) · [Liam](https://youtu.be/TvhXigW3P6Y) · [Emma](https://youtu.be/m1ZgsNWj0CI) · [Sarah Cat](https://youtu.be/_cImkjNvK44) · [Sarah Sloth Rap](https://youtu.be/hY4HORDaUgY) | youtube.com |
| Subscribers | ~250-500K est. (SocialBlade Cloudflare-blocked on direct fetch; TikTok 17.5K + Bandcamp + 1,000-video catalog + Spotify presence suggest mid-tier channel, not micro) | inferred from cross-platform footprint |

**Their weakness (our opportunity):** Visual production is mid-2010s-era — simple animations, stock character heads, basic transitions. They have NOT modernized to current YouTube thumbnail/short-form standards. The character art feels generic. There's no consistent mascot or universe — each variant (cat, sloth, alien) feels like a different channel.

---

## 2. Differentiation — How We Win

| Axis | EpicHappyBirthdays | Luna's Birthday (us) |
|---|---|---|
| **Mascot** | Generic stock characters (cat, sloth, alien heads) — no continuity | **Luna** (golden retriever puppy with star bandana — already exists in Find-Luna universe) — brand consistency across all 365 videos |
| **Visual quality** | 2015-era animation, basic text overlays | Modern Luna character + scene backgrounds (zoo, farm, beach, space) cross-sold to Find-Luna, Find-Luna-Farm, Find-Luna-Zoo |
| **Music variety** | 4-5 variant songs per name, all internally produced rock/pop | 3 song-styles (Pop, Country, Lullaby) × Luna character cameo — 9 variants per name is feasible in Year 2 |
| **Gift upsell** | Custom-song commissions via web form | Same, BUT also personalized print PDF coloring page + Luna plush mention (drives product revenue) |
| **Channel cross-pollination** | None (one-shot channel) | Funnels to findluna.smartstuffstudios.com — searches for "luna birthday" land on a channel that's adjacent to the SSS universe |
| **Production cadence** | Bursts of variants, irregular upload | 1/day metronomic schedule — algorithm favors consistency |
| **Thumbnail formula** | Plain title cards, name in standard font | Name in HUGE bubbly text + Luna mascot + bright gradient — modern CTR formula |
| **SEO on description** | Basic description | 7 related-name links, gift-share CTA, end-screen cards |
| **Music licensing** | All original / owned | Stock + MusicGen (CC-BY) — explicit license documentation per upload |

---

## 3. Year-1 Target: 365 Videos (1 Per Day)

### Cadence

- **Day 1-30:** Top 30 most-popular names (Liam, Emma, Olivia, Noah, Ava, Sophia, Isabella, Mia, Charlotte, Amelia, Oliver, Elijah, James, William, Benjamin, Lucas, Henry, Theodore, Mateo, Levi, Mason, Logan, Jackson, Ezra, Sebastian, Owen, Theodore, Daniel, Jack, Michael — picking from SSA Top Baby Names 2010-2020 list)
- **Day 31-90:** Next 60 names (mix of popular + classic + unisex)
- **Day 91-180:** Names 91-180 (uncommon but searched — Megan, Brandon, Tyler, Megan, Allison)
- **Day 181-365:** Long-tail 185 names (anything that crosses >100 searches/month on YouTube autocomplete)

### Name source

- **Primary:** US Social Security Administration Top 1000 baby names by birth year (1920-2020), deduplicated, ranked by all-time frequency.
- **Secondary:** UK ONS top baby names (for international variant).
- **Sampling:** NOT alphabetical (avoids clustering). We hit high-traffic names first so early videos drive revenue while later videos fill the long tail.

### Asset-per-video budget

| Step | Tool | Time | Cost |
|---|---|---|---|
| Pick name + write variant lines | Manual / Max (M3) | 5 min | $0 |
| Generate thumbnail | MiniMax image_generate | 30s | ~$0.02 |
| Generate song (30s clip, name embedded) | MusicGen local OR Suno OR stock | 5-30 min | $0 (MusicGen) / $0.10 (Suno) / $0 (stock) |
| Render 30s video | ffmpeg + After Effects script | 2 min | $0 |
| Upload via yt_batch_upload.py | n8n cron | 1 min | $0 |
| **Total per video** | | **~15-45 min** | **$0-$0.12** |

---

## 4. Revenue Model

### A. YouTube Ad Revenue (RPM-based)

**Two scenarios — audience setting is the biggest lever:**

| Audience setting | RPM range (2026 benchmarks) | Notes |
|---|---|---|
| **Not made for kids** (RECOMMENDED) | **$2.50–$4.50** | Per learningrevolution.net 2026 RPM data for family content; Luna is an animated puppy character, NOT a real child — defensibly non-COPPA |
| Made for kids (COPPA) | $0.10–$0.35 | pastory.app 2026: Made-for-Kids CPM is $0.25-$0.35, RPM ~10% of that |
| Sponsored integration (Year 2) | $3–$8 CPM | Kids/family sponsorships (gyre.pro 2026) |

**Recommendation: every video is uploaded as "Not made for kids."** The animation + Luna character framing makes this defensible (it's not "a child opening presents and saying their age"). The 10-20x RPM difference is the single biggest business decision.

### B. Quick math — Year 1 conservative scenario

Assumptions:
- 365 videos uploaded in 365 days
- Each video averages 1,000 views in Year 1 (Year-1 videos get traffic as people search year-round; this is conservative — EpicHB Sarah videos get 10K-100K each over their lifetime)
- Year-1 monthly views: ~30K total → ~5K/video lifetime avg in Y1
- Year-2+ compounding: each video gains another 1K-5K views/year as more people are born with or search the name

| Scenario | RPM | Year-1 revenue |
|---|---|---|
| Conservative | $2.00 | $60,000 ÷ 365 ≈ $164/video × 365 = **$60,000** |
| Wait — recalc | | 365 videos × 1,000 views × $2 RPM ÷ 1,000 = **$730 total** |

**Reality check:** The $730 figure in the task prompt is **correct as a Year-1 floor**. Birthday videos compound across years. By Year 5 the same 365 videos likely each have 5,000-10,000 views, putting annualized revenue at $5K-15K/year from a $0-$44 Year-1 investment. This is a 100-300x multi-year ROI asset.

### C. Bigger Year-2 math

If each video averages 5K lifetime views by Year 3:
- 365 × 5,000 × $2 RPM ÷ 1,000 = **$3,650/year** from the same 365-video back-catalog
- + 365 new videos × 1,000 views × $2 = **$730**
- **Year-3 total: ~$4,400/year** — passive, automated, no human work

By Year 5, with 1,825 videos and avg 8K lifetime views:
- **~$15-20K/year passive**

### D. Upsell revenue (higher-margin)

| Stream | Model | Year-1 estimate |
|---|---|---|
| Custom song (web form, like EpicHB) | $20-50/commission | $1K-5K if 50-100 commissions |
| Luna plush + song bundle (Luna's Bedtime cross-sell) | $15 wholesale + $30 retail | $2K-10K if 100-300 sales |
| Sponsored Luna videos (toy brands, kids apps) | $500-2K per video | $5K-15K if 10 sponsored |
| Spotify / Bandcamp | Per-stream royalties + album sales | $200-1K |
| **Total Year-1 upside** | | **$10K-30K** |

---

## 5. Channel Naming (Options for Tommy)

| Name | Pros | Cons |
|---|---|---|
| **Luna's Birthday** | Direct, on-brand with SSS universe, "Luna" mascot reusable | Generic — competing with EpicHB for the same term |
| **Birthday by Luna** | Conversational, character-forward | Slightly awkward |
| **Happy Birthday from Luna** | Long-tail SEO: matches "happy birthday from luna" search | Long name, harder to brand |
| **Luna Birthday Songs** | Clear, explicit, character-named | Generic |
| **Birthday Vibes by Luna** | Aligns with the existing `birthday.smartstuffstudios.com` Lane 4 subdomain | "Vibes" feels casual |

**Recommendation:** `Luna's Birthday Songs` (youtube.com/@LunasBirthdaySongs) — maximizes "luna birthday song" search and is explicit about format.

---

## 6. Music Generation — Three Paths

**Constraint:** MiniMax Music API is deprecated (per AGENTS.md / Token Efficiency Directive 2026-06-07). Each video needs ~30s of original-sounding song with the recipient's name sung in it.

### Path A: Suno (Web, manual)

- **URL:** suno.com
- **Cost:** ~$10/mo Pro plan → 500 songs, OR pay-per-credit
- **Quality:** Best in class for vocals + lyrics
- **Workflow:** Write lyrics template (30s, with [NAME] slot) → paste into Suno → generate → download MP3 → tag
- **Time per song:** 1-2 min human + 30-60s generation
- **License:** Pro plan = commercial use OK
- **Pros:** Best vocal quality, fastest path to launch
- **Cons:** Recurring cost, manual per-name generation, can't fully automate
- **Year-1 cost:** $120 (Pro subscription) for 365 songs

### Path B: AudioCraft / MusicGen (Local, free, MIT-licensed)

- **URL:** github.com/facebookresearch/audiocraft
- **Cost:** $0 — runs on local GPU (or Beacon at localhost:11434 if it has GPU access)
- **Quality:** Instrumental good, vocals limited (MusicGen is instrumental-focused; for vocals we'd need MusicGen + Bark or RVC layering)
- **Workflow:** `python -m audiocraft --model musicgen-large --text "happy birthday song for [NAME], ukulele, kids party, 30 seconds"` → .wav → tag → render
- **Time per song:** 5-15 min generation on M-series Mac, plus 30 min setup if not installed
- **License:** MIT/CC-BY (CC-BY requires attribution in description)
- **Pros:** Free, automatable, fits cron pipeline
- **Cons:** Instrumental only — no sung name. Workaround: layer a TTS voice (ElevenLabs Wiselady, per TOOLS.md) singing the name over the instrumental. Quality is "good enough for kids' birthday video" but not Suno-grade.
- **Year-1 cost:** $0 (already-eligible infrastructure)

### Path C: Stock Royalty-Free Music + Layered Name Voice

- **URL:** Pixabay Music, Uppbeat, YouTube Audio Library (free)
- **Cost:** $0
- **Quality:** Generic but professional
- **Workflow:** Download generic 30s birthday song → generate name TTS clip with ElevenLabs Wiselady → layer in Audacity/ffmpeg → tag
- **Pros:** Zero cost, fully automatable, instant
- **Cons:** All videos sound similar; brand differentiation weak
- **Year-1 cost:** $0

### Recommendation

**Start with Path A (Suno) for the first 30 videos** to nail quality and brand, then **transition to Path B (MusicGen + ElevenLabs TTS layered)** for scale. Path C is fallback.

**Per ROADMAP v3 / Token Efficiency rules:** do NOT spend on MusicGen setup until Suno proves the channel thesis. First 30 videos = paid validation; next 335 = automation.

---

## 7. Year-1 Roadmap

| Month | Milestone | Output |
|---|---|---|
| **Sep 2026** | Tommy picks name + music path; build template; first 3 test videos | 3 videos, internal review |
| **Oct 2026** | Soft launch — 30 videos (top-30 names), $0 ad spend | 30 videos, organic only |
| **Nov-Dec 2026** | Holiday push — Christmas/Snow themed variants + 60 new videos | 90 total videos |
| **Jan 2027** | First analytics review — which names performed, which formats won | Strategy reset |
| **Q1 2027** | Scale to 2/day cadence — automate MusicGen path | 270 total videos |
| **Q2 2027** | Year-1 complete — 365 videos, sponsor outreach begins | 365 videos, monetized |

---

## 8. Risks & Mitigations

| Risk | Severity | Mitigation |
|---|---|---|
| YT flags as "Made for Kids" → kills RPM | High | Set "Not made for kids" on every upload, document Luna as animated character (not real child) |
| Music licensing strike | Medium | Whitelist Suno Pro license OR CC-BY MusicGen + explicit attribution in description |
| Channel gets reported for "low quality" mass-generated content | Medium | Genuine template variation (each name gets 2-3 unique personalization lines), real Luna character art (not stock), no two videos visually identical |
| EpicHappyBirthdays sues / sends cease | Low | They're not litigious; they don't own the format. We differentiate on character + quality. |
| Suno raises prices | Low | Path B (MusicGen) is the backup, free forever |
| Cannibalization by Find-Luna-Farm / Luna's Bedtime | Low | Different content; birthday videos are short gift videos, Luna's Bedtime is podcast series |