Popiai-skill/social-media/ad-idea/references/scoring-rubric.md

# Advertising Creative Scoring Rubric & Calibration Anchors

This file defines the scoring system for advertising creative concepts, including detailed criteria for 6 scoring dimensions, calibration anchors based on real ad cases, and anti-inflation rules.

---

## 1. Six-Dimension Scoring Criteria

### Dimension 1: Strategic Fit — Weight 0.25

How well the creative concept aligns with the brief's core proposition. Not just "mentions the brand" — but "precisely addresses the communication challenge."

| Score | Criteria | Reference Case |
|-------|---------|---------------|
| 1-2 | Creative is completely unrelated to the brief; could work for any brand | — |
| 3-4 | Mentions the brand/product, but core creative is disconnected from communication goal | — |
| 5-6 | Addresses the brief's direction, but insight isn't deep or distinctive enough | Average category ads |
| 7 | Insight is precise, creative direction highly aligned with brand positioning | NetEase Yanxuan "Don't Watch This Ad" |
| 8 | Creative doesn't just respond to the brief — it redefines the communication proposition | Dove Real Beauty Sketches |
| 9 | Creative, insight, and brand exist in perfect symbiosis — remove any one and it falls apart | Nike "Dream Crazy" (Kaepernick) |
| 10 | Creative becomes part of brand equity, redefining the brand-consumer relationship | Apple "1984" |

### Dimension 2: Creative Originality — Weight 0.20

How novel the creative is. Not "never seen before" — but "seen similar things but never used this way."

| Score | Criteria | Reference Case |
|-------|---------|---------------|
| 1-2 | Completely formulaic (celebrity endorsement + feature list) | — |
| 3-4 | One or two interesting elements but overall execution is conventional | — |
| 5-6 | Clear creative concept, but not standout within the category | Most award-winning ads |
| 7 | 1-2 memorable creative elements, concept has recognizable identity | UnionPay "The Last Transfer in Tang Dynasty" |
| 8 | Pioneered an imitable creative technique | Old Spice "The Man Your Man Could Smell Like" |
| 9 | First to introduce a certain expression into advertising, sparking imitation waves | Burger King "Moldy Whopper" |
| 10 | Redefined what advertising can be | Fearless Girl |

### Dimension 3: Emotional Impact — Weight 0.20

Whether the creative can trigger strong emotional response in a short time. "3-second rule" — would you stop scrolling within the first 3 seconds?

| Score | Criteria | Reference Case |
|-------|---------|---------------|
| 1-2 | No reaction after watching, could scroll past at any time | — |
| 3-4 | "That's nice" but wouldn't watch a second time | — |
| 5-6 | One element makes you briefly pause, but overall emotional response is flat | — |
| 7 | Would casually mention to a friend "that ad was pretty interesting" | Baidu "What Did You Say?" |
| 8 | Images/copy linger in your mind after watching, want to see it again | Fotile "Love Letters in Kitchen Smoke" |
| 9 | Triggers strong emotional reaction (goosebumps / teary eyes / knowing smile), repeated viewings reveal new details | Thai Life Insurance series |
| 10 | Changes your perception of this brand (or even of advertising as a medium) | John Lewis Christmas ad series |

### Dimension 4: Scalability — Weight 0.15

Whether the creative can extend across channels, and whether the audience will proactively share.

| Score | Criteria | Reference Case |
|-------|---------|---------------|
| 1-2 | Can only be presented in a single channel, cannot be adapted | — |
| 3-4 | Can execute across 2-3 channels, but requires major rework | — |
| 5-6 | Core concept can cross channels, but creative impact diminishes when extended | — |
| 7 | Core concept has built-in extensibility, delivers different brilliant expressions per channel | Spotify Wrapped |
| 8 | Audience will proactively participate and share (UGC potential) | Ice Bucket Challenge |
| 9 | The creative itself is a sustainable platform that can generate unlimited content | Dove "Real Beauty" platform |
| 10 | Becomes a cultural phenomenon, transcending advertising itself | Nike "Just Do It" |

### Dimension 5: Feasibility — Weight 0.10

Given existing AI generation tools and reasonable production conditions, can this concept be delivered at high quality?

| Score | Criteria |
|-------|---------|
| 1-3 | Core creative exceeds current technical or budget capabilities |
| 4-5 | Some elements achievable, but key creative needs heavy post-production |
| 6-7 | Most elements can be AI-generated, minor adjustments needed |
| 8-9 | Existing MCP tools can fully deliver, with a clear tool selection plan |
| 10 | The concept itself is tailor-made for AI capabilities |

### Dimension 6: Brand Exclusivity (Brand Swap Test) — Weight 0.10

If you swap the brand for a competitor, does the concept still work? **Lower score = better** (less replaceable = more distinctive).

Note: This dimension's score is **inverted** in calculation, i.e., `actual score = 11 - raw score`.

| Raw Score | Criteria | Converted |
|-----------|---------|-----------|
| 1-2 | Completely irreplaceable; this concept belongs to this brand alone | 9-10 |
| 3-4 | Concept barely holds if you swap brands, but loses its soul | 7-8 |
| 5-6 | Swapping in a same-category brand roughly works | 5-6 |
| 7-8 | Works with any brand | 3-4 |
| 9-10 | This is a generic ad template | 1-2 |

---

## 2. Weighted Total Calculation

```
Total = Strategic Fit × 0.25
      + Creative Originality × 0.20
      + Emotional Impact × 0.20
      + Scalability × 0.15
      + Feasibility × 0.10
      + Brand Exclusivity (converted) × 0.10
```

### Score Range Meanings

| Range | Meaning | Action |
|-------|---------|--------|
| < 5.0 | Concept needs to be scrapped and restarted | Return to Phase 3, replace all methods |
| 5.0-6.4 | Has potential but problems are obvious | Target improvements on the 2 weakest dimensions |
| 6.5-7.4 | Adequate but unremarkable | Strengthen originality and emotional dimensions |
| 7.5-8.4 | Excellent, worth deep polishing | Fine-tune details, find "the last mile" |
| 8.5-9.0 | Outstanding, ready for execution | Pass — output full proposal |
| > 9.0 | Must name a real ad case it surpasses | Requires specific justification |

---

## 3. Calibration Anchors

The following real ad cases serve as scoring benchmarks, helping calibrate the absolute value of scores:

### 10-Point Level (Advertising History Milestones)

| Case | Brand | Defining Innovation |
|------|-------|-------------------|
| "1984" | Apple | One ad defined brand spirit and launched Super Bowl ad culture |
| Fearless Girl | State Street | A statue became a cultural symbol; advertising as product |
| "Just Do It" | Nike | Three words became the world's most influential brand platform, 30+ years running |

### 9-9.5 Level (Annual / Cannes Grand Prix)

| Case | Analysis |
|------|----------|
| Dove "Real Beauty Sketches" | Razor-sharp insight (female self-perception gap), simple execution that strikes the heart |
| Burger King "Moldy Whopper" | Counter-intuitive (showing a moldy burger), deep insight (proof of no preservatives) |
| Nike "Dream Crazy" (Kaepernick) | Highest-level brand attitude expression, willing to risk commercial loss for conviction |

### 8-8.5 Level (Excellent Commercial Cases)

| Case | Analysis |
|------|----------|
| Spotify Wrapped | Personalized + social currency + annual ritual, extremely strong scalability |
| Old Spice "The Man Your Man Could Smell Like" | Broke category norms, redefined men's grooming advertising with humor |
| Fotile "Love Letters in Kitchen Smoke" | Linked cooking fumes to family emotion, product as carrier |

### 7-7.5 Level (Solid Category Excellence)

| Case | Analysis |
|------|----------|
| NetEase Yanxuan "Don't Watch This Ad" | Anti-formula copy is fun, but concept has limited extensibility |
| Baidu "What Did You Say?" | Insight into elderly digital divide, strong emotional impact but slightly weak brand exclusivity |

### 5-6 Level (Mediocre)

No specific cases named — this is the zone of "celebrity + features + slogan + promotional offer."

---

## 4. Multi-Perspective Review (Perspective Panel)

For the TOP 3 concepts, simulate the following 4 roles for review:

### Creative Director Perspective
```
Review points:
- Is the concept sharp enough? Can you explain it in one sentence?
- Will it look good when executed? Is there a visual hook?
- Is there a simpler way to achieve the same effect?
- When executed with AI tools, which visuals are most likely to "break"?
```

### Strategy Director Perspective
```
Review points:
- Does the creative address the brief's core proposition?
- Does the insight hold up? Would consumers agree?
- Are brand assets being effectively leveraged?
- Have competitors done something similar? Where's our differentiation?
```

### Consumer Perspective
```
Review points:
- If I saw the first 3 seconds in my feed, would I stop scrolling?
- After watching, would I share it? What caption would I add?
- Six months later, would I still remember this ad? Which visual/copy?
- Does this ad make me want to buy the product?
```

### Cannes Jury Perspective
```
Review points:
- Among this year's international ads, does this concept stand out enough?
- Does it have a particularly strong dimension in strategy/creative/execution/technology?
- Would it inspire imitation? What core element would imitators copy?
- Does it contribute to the advertising industry?
```

---

## 5. Anti-Inflation Rules

The following 6 rules prevent inflated scores:

### Rule 1: Real Benchmark Test
> When giving 8+, you must name a real ad case at a similar score level and explain "why my concept deserves to stand alongside it."

### Rule 2: Normal Distribution Control
> In a single evaluation round, the average score of 6-9 concepts should fall between 5.5-6.5. If the average exceeds 7.0, subtract 1.5 from all scores and recalibrate.

### Rule 3: Specificity Test
> If a review uses vague adjectives like "interesting," "unique," or "innovative" without specifying *what* is interesting/unique/innovative, deduct 1 point from that dimension.

### Rule 4: "Would I Share It?" Test
> When giving Emotional Impact 8+, ask yourself: "If this were a real ad, would I proactively send it to a friend?" If you hesitate, lower the score.

### Rule 5: First Reaction Test
> What was your reaction when you first read the concept description? If it was "hmm, nice" rather than "oh!", cap at 7.

### Rule 6: Morning-After Test
> Imagine waking up the next morning and revisiting this concept — would you still think it's good? If uncertain, cap at 7.5.

---

## 6. Diagnostic Template

When a concept hasn't reached the 8.5 threshold, use this template for diagnosis:

```
## Concept {N} Diagnostic Report

### Current Score: {total}

### Strongest Dimension
{Dimension name} ({score}): {Why it's strong}

### Weakest Dimension
{Dimension name} ({score}): {Why it's weak}

### "Why isn't this a 9 yet?"
{Specific answer — vague responses not allowed}

### Improvement Directions
- Direction A: {specific action} (expected improvement: +{x} points)
- Direction B: {specific action} (expected improvement: +{x} points)

### Recommended Improvement Method
Use {method name} for Pass 2 improvement (rationale: {why this method})
```