SKL Ad Creative System · Dossier 2 of 3 · Frameworks & Rubrics

How we decide what to make — and judge what we made Two engines sit at the heart of the system. One decides where to point production when a winner lands (the Winner Iteration weighted matrix). The other judges whether any given ad is good enough to ship (the Great Ad Rubric). This dossier holds both, grounded in our own strategy docs.

Human review surface · grounded in the founder's own strategy memos (Winner Iteration Scoring Framework, the Creative Bible, the Andromeda update) · prepared by Asel · 2026-06-15 · All three dossiers live
Where this sits

The system's knowledge is organised into three dossiers. This is the second. Every abbreviation is spelled out on first use — Entity ID (Meta's label for a visually distinct ad), DTC (direct-to-camera), CTA (call to action), CPL / CPBC (cost per lead / cost per booked call).

01 The Winner Iteration weighted matrix

When an ad wins, there are several valid ways to multiply it — and no shared way to decide which to do first. This framework fixes that: it scores every iteration path and returns a ranked action list. Source: the founder's Winner Iteration Scoring Framework (March 2026).

The one idea

A winning ad's reward is not "run it harder" — it's 5+ iterations across different Entity IDs. Meta now gives near-identical ads one ticket to the auction, so the only way to scale a winner is to re-make it in genuinely distinct visual forms. The matrix decides which form to make first, by score.

The foundation — Creative Hit Rate

Before scoring paths, the system asks one question: how often do our ads actually work? That single number decides whether we spend production on finding new winners or multiplying the ones we have. Two versions are tracked:

Amelia · Portfolio Hit Rate
57% 1 in 1.8
Amelia · Tested Hit Rate (calls)
74% 14 of 19
Resulting allocation
30 / 70 new / iterate

Amelia's hit rate is exceptionally high (early-stage, low-saturation niche). Re-calculated monthly; sliced by format and by Pioneer (new concept) vs Iteration.

What the hit rate decides — production allocation

Portfolio Hit RateNew conceptsIterationWhat it means
Below 15% (1 in 7+)75–80%20–25%Winners are rare. Spend capacity finding them.
15–33%50–60%40–50%Winners at a decent rate. Balanced.
33–50%40–50%50–60%Strong. Shift toward iteration.
50%+ (Amelia: 57%)30%70%Drowning in winners. The bottleneck is multiplying them fast enough before they fatigue.

The three scoring axes

Every iteration path is scored 0–10 on three axes. The composite is their average — higher means do it first.

Entity Confidence (EC)

How sure are we this iteration reads as a new Entity ID? Cross-world or cross-style change = 10 (guaranteed). Substyle = 7. Environment-only = 5. Same person/setting, new script = 0.

Yield per Production Unit (YPU)

How many Entity IDs per unit of effort? Composite edits from existing footage score high; refilming a new day scores lower.

Speed to Ship (STS)

How fast can it be live? Composites & AI hooks 1–2 days; Skillshow 2–4 days; a refilm depends on the filming schedule.

The four paths, scored

PriorityPathECYPUSTSScore
1Composite Edit — splice hook/body/CTA across proven styles10999.3
2Skillshow Transposition — winning script as an iPad slideshow walkthrough10878.3
3AI Hook Injection (Kling) — AI-avatar 3-sec hook on a proven body8*8*8*8.0*
4Different-Style Refilm — winning script in a new validated format10646.7

*AI Hook Injection scores are provisional — they rise once the Kling workflow's output quality is validated. This is the path Gelo's 15 delivered IDA hooks sit on. Refilm scores lowest on speed but carries the highest long-term value: every new validated format compounds the footage library for all future winners.

Operating modes — composites unlock with maturity

MODE 1 · TESTING
Nothing converts yet. 100% into new concepts. No footage library worth mixing. (Unlikely for SKL given positioning.)
MODE 2 · CONVERTING
Winners exist, composites not yet unlocked. Limited format variety, so only one style has a confirmed winner. Iterate via Refilm, Skillshow, and AI Hooks. This is where Amelia is today.
MODE 3 · UNLOCKED
Composites give true leverage. Two gates met: footage in 2+ validated formats (Gate 1) and 3–5 proven winners across them (Gate 2). The full matrix opens; Composite Edit sits at the top.

Worked example — 10_IDA_ESCAPE-1.1-AccountantEscape (a confirmed winner: 16 calls, $66 CPBC, 41% hook rate)

Footage exists in four styles (Selfie, Studio, FAQ, Podcast) but only Selfie has a confirmed winner with spend behind it — so Gate 2 is unmet and composites stay blocked. We're in Mode 2. The ranked action list the matrix returns:

  • Skillshow Transposition (8.3) — top viable path; the Escape angle maps cleanly to a teaching structure. Ships within a week.
  • AI Hook Injection (8.0) — Kling hooks scripted for Bonnie; ships in 1–2 days once output is usable.
  • Different-Style Refilm (6.7) — lower score but unlocks Mode 3: if FAQ or Podcast produces a confirmed winner, composites open for the whole account.

02 The Great Ad Rubric

The matrix decides what to make. The rubric judges whether what we made is good enough to ship. It scores craft, flags the no-nos, and returns one of four verdicts — and you make the final call.

What it does

A niche-agnostic quality bar. Two layers: the no-nos (the patterns that make an ad bad — trip one and the verdict caps below green) and the quality bar (eight weighted dimensions, 0–100). Nothing auto-rejects on a market guess; the system flags, you decide.

The four verdicts

GREEN LIGHT
Ship as-is.
LIGHT TWEAK
One small fix, then ship.
RESERVE
Hold — usable later.
DISMISS
Kill it.
Grounded, and honest about its limits

Stress-tested blind against 10 real Amelia ads (5 winners, 5 losers): it caught 4 of 5 losers and the 3 lowest scores were all genuine losers — a strong filter against bad ideas. It over-rejected some winners, which drove two design choices (proof is a scored deduction not a kill-switch; format-execution fit carries real 15% weight). Weights stay Provisional until more performance data lands.

→ Open the full Great Ad Rubric (Consolidated V1) — the eight dimensions, the no-nos, the worked example, the back-test, and the glossary.

03 The frameworks underneath

Both engines stand on two foundational strategy memos. They're the "why" behind the scores.

The Creative Bible — Andromeda-era multiplication

Meta's Andromeda delivery system clusters ads by visual similarity and gives each cluster one ticket to the auction. Above ~60% similarity, ads compete with each other instead of reaching new audiences. So the unit of scale is the Entity ID, not the ad. The Bible's multiplication matrix — 6 creative worlds × 8 styles — turns one proven concept into 19–22 distinct Entity IDs. AI-generated hooks are World 6: 10–15 fresh-face hook variations in an afternoon, spliced onto a proven body.

The Andromeda update — minority hooks & simple structure

From the Jeremy Haynes mastermind: the old game of one "majority hook" milked across 100 iterations is dead. The new game is 25–50 genuinely distinct ads, each speaking to a specific "minority hook" — the algorithm matches each message to its small, relevant audience pocket. Campaign structure stays simple (one cold campaign, a few ad sets, 25–30 ads each); fatigue is managed with the 50/50 test then a full refresh.

Agent-facing versions of these live in standards/creative-strategy-framework (hit-rate methodology, path scoring) and standards/style-system (14 format profiles with Entity-ID signal classification). This dossier is the human-facing read.

04 How the two engines work together

A WINNER LANDS
Confirmed by spend + Hyros attribution (e.g. AccountantEscape).
MATRIX DECIDES
Hit rate sets the new/iterate split; the four paths are scored; a ranked action list comes out (Composite → Skillshow → AI Hook → Refilm).
RUBRIC JUDGES
Each new iteration is scored before it ships — craft bar + no-nos → one of four verdicts. You make the final call.
LOG & LEARN
Shipped ads feed performance back; hit rate and rubric weights re-calculate monthly. The system sharpens on real data, never auto-re-weights.