skills/creative-director/references/legendary-patterns.md

23.6 KB · Sep 30, 2026 · 23:14 UTC

# Legendary Pattern Map — P01–P18

Companion to `legendary-campaigns/MOC-pattern.md`. That file lists canonical cases per pattern; this file defines the mechanics, rationale, and stealing strategy for each.

**What this is.** A calibration instrument. Every strong idea maps to one or two of these 18 patterns. Knowing the pattern tells you (a) whether the mechanic is proven, (b) how saturated the space is, and (c) what the originality challenge is.

**How to use for calibration.** When an idea is in a pattern with 50+ canonical cases in the library, originality pressure increases — the mechanic is proven but crowded. The question shifts from "will it work?" to "what makes this version necessary?" See Pattern Calibration Workflow below.

**How to use for pre-mortem.** After selecting a finalist concept, run it through the Pre-Mortem Template at the bottom of this file before presenting. The template surfaces failure modes before the client does.

---

## Pattern Map

---

## P01 — Format as Idea

**Core mechanic:** The medium or platform constraint becomes the concept itself. The ad cannot exist in any other format without losing its meaning.

**Why it works:** Novelty is structural, not decorative. The audience cannot separate "the idea" from "the container," which produces surprise and memorability without needing exceptional content.

**Canonical examples:**
- IKEA "Pee Ad" — pregnancy test embedded in a magazine spread; positive result reveals a crib discount
- Burger King "Whopper Detour" — 1-cent order only activates within 200m of a McDonald's; geolocation is the mechanic
- Stella Artois "Eulogies" — beer ad formatted as an obituary for an unfinished pint

**Stealing strategy:** Audit the format's native constraints (size, timing, interaction, context of consumption). Find the one constraint that most brands treat as a limitation. Make it load-bearing.

**Saturation:** 6 cases in library (5 fully documented).

---

## P02 — Enemy or Conflict

**Core mechanic:** The brand builds its communication around an opponent — a competitor, a habit, a stereotype, or a broken system. The conflict simplifies the message and consolidates audience allegiance.

**Why it works:** Humans organize identity around opposition. A clear enemy makes the brand's position immediately legible and gives the audience a side to take.

**Canonical examples:**
- Apple "Get a Mac" — direct human personification of Mac vs. PC
- Burger King "Moldy Whopper" — rotting burger as attack on preservatives across the category
- Greenpeace vs. Volkswagen "The Dark Side" — ecological attack via Star Wars parody

**Stealing strategy:** Name the enemy precisely. "The category" is too vague. "The belief that X is inevitable" is a target. The more specific the opponent, the sharper the brand position.

**Saturation:** 15 cases in library (7 fully documented).

---

## P03 — Behavior Inversion

**Core mechanic:** Identify an arbitrary category rule — something everyone does without questioning — and violate it literally. The longer the rule has held, the louder the disruption.

**Why it works:** Category codes are learned, not natural. Breaking them signals confidence and reframes what the brand represents. The audience recognizes the violation immediately because they have internalized the rule.

**Canonical examples:**
- Dove "Real Beauty" — beauty brand casts real women instead of models
- Intermarche "Inglorious Fruits and Vegetables" — retailer sells visually imperfect produce at a discount, inverting quality signaling
- Libresse "Bloodnormal" — feminine care shows menstrual blood as red, not blue

**Stealing strategy:** List every visual and tonal convention the category uses without question. Pick the one most taken for granted. Invert it without softening.

**Saturation:** 44 cases in library (21 fully documented). High saturation — originality requires a sharper or more unexpected inversion target.

---

## P04 — Brand as Activist

**Core mechanic:** The brand takes a position on a social, environmental, or political issue and acts on it at cost. The action, not the statement, is the campaign.

**Why it works:** Stated values are cheap; costly actions are credible. When a brand sacrifices something (revenue, audience, contracts), it demonstrates genuine alignment with its stated beliefs.

**Canonical examples:**
- Patagonia "Don't Buy This Jacket" — Black Friday anti-consumption ad by a consumer goods brand
- Nike "Dream Crazy" — support for Kaepernick at the peak of his controversy
- Tony's Chocolonely — unequal chocolate bar segments as physical embodiment of unequal cocoa labor conditions

**Stealing strategy:** Find the issue the brand's actual business practices can credibly support. Avoid issues where the brand has no authentic stake. The action must cost something visible.

**Saturation:** 28 cases in library (14 fully documented).

---

## P05 — Cultural Hijack

**Core mechanic:** The brand inserts itself into an already-live cultural event, news cycle, or meme. Speed is the primary asset; the cultural context provides the meaning.

**Why it works:** Attention is already concentrated on the event. The brand borrows that attention at marginal cost — but only if the insertion is fast, appropriate, and adds something rather than merely referencing.

**Canonical examples:**
- Oreo "Dunk in the Dark" — tweet during Super Bowl XLVII blackout, posted within minutes
- KFC "FCK" — full-page apology for chicken shortage with scrambled logo letters
- IHOP to IHOb — week-long name change generating organic media coverage around a menu pivot

**Stealing strategy:** Build a reactive capability before the cultural moment, not during it. Have approved concepts for predictable events (annual moments, category flashpoints). The trap is speed without relevance.

**Saturation:** 35 cases in library (9 fully documented).

---

## P06 — Limitation as Power

**Core mechanic:** A genuine weakness, small scale, or structural constraint is reframed as the brand's defining advantage. The limitation becomes the positioning.

**Why it works:** Admitting a weakness signals honesty and confidence simultaneously. It preempts the competitor's attack and reframes the audience's evaluation criteria in the brand's favor.

**Canonical examples:**
- Avis "We Try Harder" — being #2 to Hertz becomes a service promise
- VW "Think Small" — the Beetle's smallness positioned against American excess
- Buckley's "It Tastes Awful. And It Works." — the product's worst attribute becomes its efficacy proof

**Stealing strategy:** Find what the brand genuinely cannot compete on. Turn that into the claim. The limitation must be real — invented weakness reads as false modesty.

**Saturation:** 9 cases in library (6 fully documented).

---

## P07 — Invisible Brand

**Core mechanic:** Minimal logo, minimal product presence. The brand trusts its cultural recognition enough to let the narrative carry the full weight.

**Why it works:** Restraint signals confidence. Audiences are sophisticated enough to recognize excessive branding as insecurity. When the brand disappears, the story expands.

**Canonical examples:**
- Coca-Cola "Hilltop" — brand identity conveyed entirely through mood and music
- Apple "1984" — product visible for two seconds in the closing frame
- Guinness "Surfer" — three-minute wave narrative; the pint appears at the end

**Stealing strategy:** Ask what percentage of the ad could be removed without losing the brand signal. If the answer is "none," the brand is not confident enough. Invisible brand only works when the brand is already legible through tone, casting, or visual language.

**Saturation:** 15 cases in library (7 fully documented).

---

## P08 — Craft as Message

**Core mechanic:** The quality of execution is itself the argument. The way the ad is made demonstrates the brand's commitment to excellence without stating it.

**Why it works:** Showing is more credible than telling. When the craft is demonstrably extraordinary — scale, precision, duration, physical reality — it transfers to product perception without a single claim.

**Canonical examples:**
- Sony "Bravia Balls" — 250,000 colored balls released down San Francisco streets, filmed for real
- Cadbury "Gorilla" — gorilla performing Phil Collins on drums, no product logic, pure tonal perfection
- Chanel No. 5 "The Film" — three-minute Baz Luhrmann cinematic narrative with Nicole Kidman

**Stealing strategy:** Find the one execution choice that cannot be faked or cheapened. Build the budget around that element. Everything else is secondary.

**Saturation:** 49 cases in library (5 fully documented). High saturation — execution must be genuinely exceptional, not just expensive.

---

## P09 — User as Co-Author

**Core mechanic:** The audience becomes the content. The campaign provides a mechanic; users generate the actual media that carries the message.

**Why it works:** Participation creates ownership. When people make something, they share it. The brand's job is to design the mechanic, not the content. UGC volume is structurally impossible to replicate with paid media.

**Canonical examples:**
- ALS Ice Bucket Challenge — users create and nominate, brand (awareness) scales without brand spend
- Spotify Wrapped — personal listening data becomes shareable social content
- Share a Coke — name-printed bottles turn every purchase into a personal artifact

**Stealing strategy:** Design the input/output loop. The mechanic must be simple enough to execute in under 60 seconds and produce an output the user wants to keep or share independently of brand loyalty.

**Saturation:** 100 cases in library (26 fully documented). Very high saturation — the mechanic must have a genuinely novel trigger or output format.

---

## P10 — Serialization and Ritual

**Core mechanic:** The campaign becomes a repeatable structure — a recurring character, annual event, or seasonal tradition. Repetition over time is the medium.

**Why it works:** Ritual creates anticipation. When an audience expects something and it arrives on schedule, the emotional response is amplified by the waiting. The brand occupies calendar space rather than media space.

**Canonical examples:**
- Coca-Cola Christmas Truck — Santa in a red truck each December, now a cultural calendar marker in multiple markets
- Old Spice "The Man Your Man Could Smell Like" — series with Isaiah Mustafa, including real-time celebrity response videos
- John Lewis Christmas — annual UK emotional film that audiences actively wait for each December

**Stealing strategy:** Build the tradition first, not the ad. Define the cadence, the invariant element (character, format, date), and the variable (story). The invariant must be ownable.

**Saturation:** 30 cases in library (10 fully documented).

---

## P11 — Absurd as Carrier

**Core mechanic:** The idea is delivered via pure absurdism, anti-comedy, or surreal logic. The audience's rational defenses are bypassed; the brand message travels through the gap.

**Why it works:** Absurd content is inherently shareable because it cannot be explained, only experienced. It also differentiates sharply in a category with earnest competitors. The brand appears confident enough to not explain itself.

**Canonical examples:**
- Skittles "Touch the Rainbow" — man transforms objects into Skittles on contact; no narrative logic applied
- Old Spice "The Man Your Man Could Smell Like" — horse on a beach, diamonds, escalating impossibility
- Cadbury "Eyebrows" — children moving eyebrows to electronica; no product justification offered

**Stealing strategy:** Commit completely. Half-absurdist work reads as confused. Choose a single absurd premise and follow its internal logic without breaking to explain. The moment the brand explains the joke, the absurd collapses.

**Saturation:** 239 cases in library (22 fully documented). Extremely high saturation — requires a genuinely distinct absurd premise, not just "weird."

---

## P12 — Social Experiment

**Core mechanic:** The brand constructs a real situation where real people reveal the insight or problem. The audience receives the result as evidence, not as a brand claim.

**Why it works:** Third-party credibility without third-party cost. When people see other people react authentically, they believe it in a way they cannot believe a scripted claim. The brand is the designer of the experiment, not the advocate.

**Canonical examples:**
- Dove "Real Beauty Sketches" — FBI composite artist draws women from self-description vs. stranger description
- Always "Like a Girl" — same action performed differently by children vs. older subjects
- TV2 Denmark "All That We Share" — people sorted by stereotype, then resorted by shared experience

**Stealing strategy:** The experiment must be genuinely rigged to reveal the insight, not to confirm a predetermined conclusion. If the result could plausibly go the other way, the experiment is credible. If it cannot, it reads as manufactured.

**Saturation:** 9 cases fully documented in library.

---

## P13 — Truth Telling

**Core mechanic:** A powerful human or social truth is stated without softening, in a category that habitually avoids it. Directness is the differentiator.

**Why it works:** Most categories develop shared conventions for what is not said. The brand that says it earns disproportionate trust and attention precisely because the convention exists. The contrast is the mechanism.

**Canonical examples:**
- Sandy Hook Promise "Evan" — school romance narrative that reveals a school shooting in the final frames
- P&G "The Talk" — Black mothers across decades preparing children for encounters with police
- TAC "Meet Graham" — the only human body anatomically capable of surviving a car crash, presented as a public sculpture

**Stealing strategy:** List what the category systematically avoids saying. Find the one avoidance that is most recognized and least addressed. Say it with precision, not with drama — the truth does not need decoration.

**Saturation:** 87 cases in library (34 fully documented). High saturation — the truth being told must be genuinely unspoken, not simply stated more bluntly.

---

## P14 — Product as Proof

**Core mechanic:** The campaign physically demonstrates the product claim rather than asserting it. The demonstration must be impossible to fake or recreate without the product's actual performance.

**Why it works:** Claims are forgettable; demonstrations are memorable and credible. When the brand stakes the entire execution on a product performing, the audience infers that performance is real.

**Canonical examples:**
- Volvo Trucks "Epic Split" — Van Damme performs full lateral splits between two moving trucks to demonstrate precision steering
- Honda "The Cog" — Rube Goldberg machine built entirely from Honda Accord parts, single take, 606 attempts
- Apple "Shot on iPhone" — consumer photography and video becomes the global outdoor campaign

**Stealing strategy:** Find the product attribute that can be made physically undeniable. The demonstration must be something the audience cannot attribute to camera tricks, editing, or coincidence. If it can be doubted, it loses.

**Saturation:** 17 cases in library (5 fully documented).

---

## P15 — Benefit Hyperbole

**Core mechanic:** The product's benefit is taken to a physical, fantastical, or extreme literal conclusion. Exaggeration makes an abstract promise visible and memorable.

**Why it works:** Abstract benefits (energy, protection, confidence) are not differentiating because they cannot be compared. Hyperbole materializes the abstract into a specific, vivid image that the audience can hold.

**Canonical examples:**
- Red Bull "Stratos" — Felix Baumgartner jumps from the stratosphere as literal embodiment of "gives you wings"
- Nike "Breaking2" — attempt to run a sub-2-hour marathon as demonstration of human potential
- Old Spice "The Man Your Man Could Smell Like" — impossible lifestyle cascade as hyperbolic product effect

**Stealing strategy:** Take the core benefit and follow it to its most extreme logical conclusion. The hyperbole must be internally consistent — the audience needs to understand the exaggeration is intentional and confident, not accidental.

**Saturation:** 6 cases fully documented in library.

---

## P16 — Long-form Drama

**Core mechanic:** A cinematic narrative in extended format. Emotion accumulates through story structure — conflict, loss, choice, or reconciliation — rather than through a single hook.

**Why it works:** Long-form earns disproportionate attention when the story is genuinely worth the time. It creates a quality signal (the brand paid for something substantive) and allows for emotional complexity that 30-second formats cannot hold.

**Canonical examples:**
- John Lewis "Man on the Moon" — elderly man alone on the moon; a child sends him a telescope via a gift
- Sainsbury's "1914" — Christmas truce in World War I trenches, told as a feature-quality period piece
- P&G "Thank You, Mom" — Olympic athletes' journeys framed through their mothers' sacrifices

**Stealing strategy:** The story must justify the length. If it can be told in 90 seconds without loss, it should be. The additional time must unlock emotional complexity or narrative turns that are structurally impossible in short format.

**Saturation:** 261 cases in library (43 fully documented). Extremely high saturation — the story must be genuinely exceptional in originality or execution, not simply well-produced.

---

## P17 — Design as Idea

**Core mechanic:** Typography, packaging, or visual design carries the entire concept. Minimal copy; maximum form. The design decision is the message.

**Why it works:** Pure visual ideas are faster than verbal ones and cross language barriers. When design does the heavy lifting, the idea is often simpler and more durable than any headline could make it.

**Canonical examples:**
- The Economist — white text on red, no image, no explanation; confidence through restraint
- Heinz "Tilt" — label applied at the angle you'd hold the bottle to pour; instruction as design
- WWF "50% / 50%" — half panda head, half negative space; the missing half communicates extinction

**Stealing strategy:** Remove all copy. Does the idea survive? If it requires explanation, it is not yet a design idea — it is a design execution of a verbal idea. Keep reducing until the remaining form is self-sufficient.

**Saturation:** 72 cases in library (23 fully documented).

---

## P18 — Tech as Canvas

**Core mechanic:** A new technology or platform is the idea itself. The technology is not a distribution channel; it is the demonstration of what the brand believes the future looks like.

**Why it works:** New technology carries intrinsic attention. When a brand is first to use a technology in a meaningful (not gimmicky) way, it earns an association with that technology's promise. The medium argues for the brand's modernity without a single claim.

**Canonical examples:**
- Burger King "Burn That Ad" — AR app burns competitor outdoor ads and delivers a coupon in their place
- Snickers "Hungerithm" — algorithm tracks internet sentiment and prices Snickers inversely to collective online anger
- Spotify Wrapped — annual personalized infographic that invented a new genre of shareable brand content

**Stealing strategy:** The technology must do something the idea requires, not something the idea merely tolerates. If you could remove the technology and keep the idea, the technology is decoration. The question is: what becomes possible for the audience only because of this technology?

**Saturation:** 33 cases in library (11 fully documented).

---

## Pre-Mortem

**Premise.** Imagine the campaign launched six months ago and failed catastrophically. Why?

Run this on the finalist concept before client presentation. Not as a ritual — as a filter.

### Process

1. Lock the concept. Do not run pre-mortem on ideas still in development.
2. Each participant independently generates failure reasons (5 minutes, no discussion).
3. Consolidate into a priority table (below).
4. For each High × High cell: define a mitigation or kill the concept.

### Five Standard Failure Modes to Check

| Failure Mode | Question to Ask |
|---|---|
| Brand misalignment | Does the brand have authentic standing to do this? Would a skeptic find a gap between claim and behavior? |
| Cultural backlash | Which group could legitimately feel harmed, mocked, or excluded? What is the realistic media headline in the worst case? |
| Execution flaw | Is there a production dependency that, if it fails, collapses the entire idea? What is the single-point-of-failure? |
| Metric mismatch | Does this campaign produce the metrics the client is accountable for? Or does it produce different metrics the client cannot report? |
| Saturation | Has someone already done this, or a close enough version? When did they do it, and in what market? |

### Probability x Impact Matrix

Classify each identified failure mode:

### Output Format

After mapping, produce a priority table sorted by Probability x Impact (descending):

| # | Failure mode | Probability | Impact | Mitigation |
|---|---|---|---|---|
| 1 | [description] | H/M/L | H/M/L | [action] or KILL |
| 2 | ... | | | |

**Verdict options:**
- Go — all High×High cells have viable mitigations
- Revise — specific changes required before presenting (list them)
- Kill — at least one High×High cell has no viable mitigation

---

## Pattern Calibration Workflow

Use this sequence after generating a candidate idea:

1. **Identify the pattern.** Which P-code does this idea's core mechanic map to? If it maps to none, it is either genuinely novel or not yet a mechanic — distinguish between the two.

2. **Check the library.** Open `legendary-campaigns/MOC-pattern.md` and locate the canonical cases under that pattern. Read the mechanic descriptions, not just the names.

3. **Differentiation check.** Compare your idea's mechanic to the canonical cases. Articulate the specific difference in one sentence. "We are doing X, which differs from prior cases because Y." If you cannot complete that sentence, the idea is not differentiated.

4. **Saturation adjustment.** If the library shows 50+ cases in the pattern, set the originality ceiling at 7/10 maximum unless the differentiation check yields a demonstrably novel mechanic. Saturated patterns require higher execution standards to compensate.

5. **Record the pattern tag** in the idea brief as `pattern: P{N}` for tracking and future reference.

---

## Use in creative-director Skill

- **Phase 4 calibration block.** After generating candidate ideas, run each through this pattern map. Identify the P-code, check saturation, and surface the originality challenge before scoring. A high-execution idea in a saturated pattern scores differently than the same execution in an undersaturated one.

- **Pass 4-5 Pre-Mortem.** Before the finalist concept reaches the client presentation, apply the Pre-Mortem Template. Output the priority table and attach the Verdict (Go / Revise / Kill) to the concept record. If the verdict is Revise or Kill, loop back to ideation before proceeding.

---

*Author: Serge Shima ([t.me/aimastersme](https://t.me/aimastersme) · [sergeshima.com](https://sergeshima.com) · [aimasters.me](https://aimasters.me)) · License: CC BY 4.0 — attribution required · Source: [smixs/creative-director-skill](https://github.com/smixs/creative-director-skill)*

SHA-256: 333c0abd69a066bd440ef3560e195e2993b816508515841cce7531681d388e08