The big idea is Ogilvy's term for the one thought that carries a campaign for years: Think Small, We Try Harder, Got Milk?, Just Do It. This skill finds big ideas by method. The first competent idea is the category average, the one every rival agency also brings. The method is built to get past it.
The method comes from James Webb Young: an idea is a new combination of old elements, and the skill of combining is the skill of seeing relationships between facts. So the work is raw material in, many combinations out, then ruthless judgment. Most weak ad ideas fail at the first stage, from too little raw material, rather than at the combining.
Vocabulary
These words are used precisely throughout the skill and its references.
| Word | Meaning |
|---|---|
| Raw material | Specific facts about the product, truths about the audience, the category's conventions, and the culture around them |
| Insight | A truth the audience recognizes the moment they hear it, which nobody in the category has said |
| Proposition | The single thing the audience should believe or feel afterward. One thought, single-minded |
| Idea | The surprising, ownable form that dramatizes the proposition. Fits in one sentence and generates many executions |
| Execution | One concrete piece: a print ad, a film, a stunt, a promo mechanic, a headline |
| Lens | A generative move that turns raw material into ideas, drawn from the canon (Lenses) |
| Ancestor | A landmark campaign that solved a similar problem with a similar move (Cases) |
| Wallpaper | What every brand in the category says and shows. The idea breaks it |
Pick the path
| Request | Path |
|---|---|
| Campaign, ad concept, brand platform | Full workflow, three concepts |
| Promotion, stunt, launch, activation | Full workflow; executions are mechanics (who does what, what they get, why they tell people) |
| Tagline, headline, slogan, subject line, name | Steps 1–2 lightly, step 4 with the copy lenses, 40+ lines written, best 8–12 delivered grouped by angle |
| Copy for an existing idea | Read Executions and write |
| Non-ad writing that needs an angle (launch post, talk title, pitch) | Same method; the proposition is what the reader should take away |
Workflow
1. Brief
Extract the product, audience, business problem, channel and budget signals, tone, and mandatories. Then restate the problem in one sentence that begins "The problem is that people…". The stated brief often hides the real problem; Dave Trott's predatory thinking starts by reframing it, and Rory Sutherland asks whether it is a problem of perception rather than of product.
Ask questions only when a missing answer changes what kind of idea is needed (for example, whether the goal is trial or loyalty). Otherwise state the assumption and continue.
2. Raw material
This is Young's first step, and the one most often skipped. Build a fact sheet:
- Product: at least 12 specific facts. How it is made, what is in it, numbers, origin, the founder's obsessions, the odd details, what it does that nobody mentions. Claude Hopkins sold Schlitz on steam-cleaned bottles and artesian wells, a process every brewer used and none had told the public about. Ogilvy found the Rolls-Royce clock in an engineering report. When the product is real, search the web for facts and cite what you find.
- Audience: at least 5 truths. Behaviors, small rituals, tensions, the thing they would never say out loud, the moment the product matters most.
- Wallpaper: the category's conventions. The claims, images, tone, and colors every competitor uses.
- Culture: 2–3 live currents the brand could join or push against.
Done when each product fact is specific enough that a competitor could not claim it word for word, or is marked as category-generic (generic facts can still win, as preemptive claims).
3. Strategy
Write three distinct strategic angles, each as an insight plus a proposition. Make them differ in strategy, not wording: one from a product fact (USP or inherent drama), one from an audience tension, and one from positioning against the category or a rival (an enemy, a flaw turned into a virtue, a word to own). The canon behind each move is in Canon. Carry the two or three strongest forward.
4. Generate wide
Run every lens in Lenses against the fact sheet and the angles. Write each idea as one line, fast and without judging. Aim for 40 or more raw ideas; quantity is the method here. Then force combinations: pair an unrelated product fact with a lens, or an audience truth with a case's move, and write what comes out.
Mark the first five or so ideas you wrote as the baseline: the obvious answers. They stay in the pile only if they survive step 6 against everything else.
5. The rival's pitch
Young's step three is incubation, setting the problem aside so the mind can recombine it. The agent version: write the three concepts a competent rival agency would bring to this brief. Strike your own ideas that resemble them. What remains is unexpected by definition.
6. Select
Score the survivors against the tests below and keep three concepts that answer different strategic angles.
| Test | Passes when |
|---|---|
| One sentence | The idea fits in one sentence a stranger would repeat |
| Ownable | Swap in a competitor's name and it breaks |
| Gasp | Ogilvy's questions: did it make me gasp, and do I wish I had thought of it? |
| On strategy | It dramatizes the proposition; the audience walks away believing it |
| Branded | The product is the hero of the idea and not a sign-off, so nobody remembers the ad and forgets the brand |
| Legs | It yields three or more executions across different media; could it run for years? |
| Fame | People would retell it unprompted (Feldwick's showmanship, Berger's social currency) |
| Truth | The product lives up to it. Bernbach: a great campaign makes a bad product fail faster |
7. Develop
Build each concept to the output format below. Write the hero execution in enough detail to see or hear it: the actual headline and visual, the film's story in a few sentences, the stunt's mechanics from first touch to the moment someone shares it. Write the copy to Executions. Name an ancestor from Cases for each concept and say what this idea takes from it and where it departs.
Done when each concept has an insight, a proposition, a one-sentence idea, a hero execution, three or more extensions, and an honest risk, and the three concepts would lead to three different campaigns.
Output
# [Brand / product]: concepts
## The problem
[One sentence: "The problem is that people…"] plus any assumptions.
## Raw material
[The fact sheet, condensed to the facts the concepts use, plus the category wallpaper.]
## Concept 1: [Name]
**The idea:** [One sentence.]
**Insight:** [...]
**Proposition:** [...]
**Ancestor:** [Campaign]: what this takes from it, and where it departs.
**Hero execution:** [Headline and visual / film story / stunt mechanics, written out.]
**Extensions:** [3+ executions across media: print or OOH, film, social, promotion, retail, PR.]
**Lines:** [2–4 tagline or headline options.]
**Why it works:** [The principle at work, in a sentence or two.]
**Risk:** [What could go wrong, and what would have to be true for it to work.]
## Concept 2 … Concept 3
## Recommendation
[The one you would take into the room, and why.]
## Also ran
[Strong ideas cut in selection, one line each.]
For tagline and headline requests, replace the concepts with lines grouped by angle, each group headed by its insight, with your recommendation first.
Related skills
video-scriptturns a chosen concept's film into a shootable script.creative-directorbuilds the visual identity around a brand.clean-writingtightens long copy.
Reference documents
| Reference | Use during |
|---|---|
| Canon | Steps 1, 3, and 6: what each master taught, as a principle you can apply and a question to ask |
| Lenses | Step 4: the generative moves, each with its prompt and a campaign that used it |
| Cases | Steps 4 and 7: landmark campaigns broken into problem, insight, leap, and why it worked |
| Executions | Step 7: headlines, body copy, taglines, print and outdoor, film, promotions and stunts, social |