What the data actually says
Every number below comes from the channel's YouTube Analytics API, pulled 10 July 2026.
Organic is tiny — but it is the right kind of tiny.
97% of views are paid. Yet the organic viewers that do arrive are the deepest watchers on the channel: search viewers average 8+ minutes, playlist listeners ~3 minutes per view. The raw material for organic growth is quality, not quantity — the job is to multiply it.
Traffic, last 28 days (engaged views)
| Source | Views | Engaged | Watch min |
|---|---|---|---|
| Paid advertising | 3,461 | 1,518 | 3,101 |
| Suggested videos | 34 | 34 | 47 |
| Playlists | 34 | 34 | 94 |
| Subscribers | 33 | 33 | 55 |
| Search | 7 | 7 | 58 |
| Shorts feed | ~207 raw plays | ≈0 | — |
The Shorts row is the tell: the feed is willing to test the content on strangers (~207 plays), and viewers bail at ~18 seconds of ~34. Distribution exists; retention doesn't — yet.
Retention identifies the product (lifetime)
| Track | Avg watched | Held |
|---|---|---|
| We Don't Break, We Reload | 2:07 of 3:37 | 58.6% |
| Welcome To The Crest | 1:50 of 3:12 | 57.4% |
| Built Different | 1:47 of 3:22 | 53.2% · most viewed (1,782) |
| Pressure Makes Warriors | 1:06 of 3:23 | 33% |
| Loose Cannons Live | 0:41 of 2:28 | 27.8% |
| Underdog | too new — analytics lag; strongest paid sub-converter (~29p/sub) | |
The three engines
One engine per goal: views, site engagement, subs — each powered by something the data already proved.
Engine 1 Match-day moment engine → organic views + subs
One 20–30s chant Short per fixture, posted ~3 hours before kickoff; a second cut ready for full-time — win gets the celebration edit, loss gets the defiance edit (We Don't Break, We Reload is literally built for bad days).
The Shorts formula that fixes the 18-second bail: drop lands inside 1 second → karaoke-style chant lyrics on screen (participation = rewatch) → beat switch-up at 0:12–0:15 → seamless loop. Loops push retention past 100%, the strongest Shorts signal there is.
Engine 2 Searchable canon → organic views + website
Search viewers are the deepest watchers; make the catalogue findable and the site the destination. Descriptive retitling (policy-safe, no artist-name games), full lyrics in every description (lyric searches are how chants get learned), and chant sheets on the website — one page per anthem with lyrics, printable PDF and the embedded video. The chant sheet is the shareable object: the thing that gets forwarded into group chats before a match. Every video pins a comment linking its sheet: “Learn it before Saturday →”. End screens all point at the Anthems playlist (3 min/view, the engagement champion).
Engine 3 Community flywheel → subs + engagement
At 102 subs, every fan can be treated like a founding member. Reply to every comment within hours (the 8am daily check flags them). Weekly community-tab polls — “which anthem gets the next video?” — because voters become invested, and invested viewers subscribe. Fan-space participation done honestly: contribute where music is on-topic, never link-spam; good chant sheets travel on their own.
What stays paid (small, deliberate)
Promos aren't the enemy — dependency is. Keep the two proven sub-converters (Underdog, Loose Cannons) at low spend as a floor while the engines spin up; the rest stays off per the promo analysis.
Storyboards — batch one
All Shorts: 9:16, drop inside 1s, karaoke lyrics, switch-up ~0:12–0:15, seamless loop. Audio is always the actual track — that's the product. Visuals cut from the existing music videos first; AI B-roll only to fill gaps, each clip approved before use.
S1 · “We Don't Break — Matchday Cut” (25s)
| Time | Visual | Text on screen |
|---|---|---|
| 0:00–0:01 | Hard cut: red smoke bursts at lens; drop hits immediately | — |
| 0:01–0:04 | Crowd silhouettes, fists in rhythm | “NORTH LONDON, STAND UP” |
| 0:04–0:12 | Street march, floodlights | Karaoke chant lyrics, word-by-word |
| 0:12–0:15 | Switch-up: beat flip + crowd chant layer | “LOUDER.” |
| 0:15–0:22 | Fast cuts: scarf, cannon emblem, smoke | Chant lyrics continue |
| 0:22–0:25 | Frame matching 0:00 → seamless loop | “Full anthem → @RogueCanon” (final 2s) |
S2 · “Underdog — Learn the Chant” (30s)
Call-and-response: line appears → beat gap for the viewer to answer → next line. End card: “Can you keep up? Full chant in the comments.” Pinned comment carries the full lyrics + chant-sheet link. Engineered for comments — engagement bait that's actually participation.
S3 · “Crest — 15-Second Drop”
Zero text until the final frame. Pure drop + the strongest visual run from the Crest video. Final frame: “Welcome to the Crest →”. This is the A/B test: does raw sound beat the lyric formats on retention?
S4 · Community poll
“Next anthem gets the full video treatment. You pick:” — three options plus “comment your own”. The result gets pinned in the winner's premiere.
S5 · Chant sheets (website, ×6)
One page per anthem: large-type lyrics, printable PDF button, embedded video, playlist link. Site nav gains “CHANTS”. Built on PageMotor.
S6 · Metadata pass (all six public tracks)
New descriptions: hook line → full lyrics → chant-sheet link → playlist link. Titles get descriptive suffixes where honest (e.g. “— North London Terrace Anthem”). No artist-name stuffing — that's a YouTube policy trap, and we don't need it.
Targets & measurement
Tracked automatically by the 8am daily check; reviewed weekly.
| Goal | Baseline (10 Jul) | Target |
|---|---|---|
| Organic subscribers | ~1/day, paid-assisted | +25 organic in first 30 days, compounding after |
| Engaged organic views | ~128 / 28 days | 500+ / 28 days by end of August (match-day engine online) |
| Website engagement | ~0 tracked organic visits | 6 chant sheets live + first 100 organic visits |
Kill/scale rule, same as promos: any Short format holding under 40% retention twice gets reworked; anything over 60% becomes a series. The next real milestone behind all three: 1,000 subs + 4,000 watch hours = monetisation.
Execution
Split by who can actually do what.
Kenn: approve this playbook
Everything below fires on the green light.
Kenn: supply the lyrics
All six anthems — paste them in chat or point at files. They feed the descriptions (S6) and the chant sheets (S5).
Claude: metadata pass + chant sheets
Descriptions and titles updated via the YouTube API; six chant-sheet pages built on PageMotor with printable PDFs; pinned-comment text prepared per video (pinning is a 5-second manual step — the API can't pin).
Kenn: cut S1–S3
From existing music-video footage, per the storyboards. Karaoke text in your editor; timing beats polish.
Both: calendar + post
Claude pulls the fixture list and builds the posting calendar (Shorts ~3h before kickoff, full-time cut on the whistle, one per day max). Kenn posts; the daily check scores every Short's retention and flags what to double down on.