A typical brainstorm meeting produces three variants of the same safe idea, then everyone agrees the middle one is “the strongest direction.” Real differentiation comes from forcing genuinely different stances and then scoring them honestly — not from a louder whiteboard. This walks through a 5-step AI workflow that produces 12 differentiated campaign concepts in 30 minutes, ready to bring to a stakeholder review.
What this covers
A 5-step ideation workflow that turns a one-paragraph brief into 12 differentiated campaign concepts — each with a headline, hook, lead channel, and execution idea, then scored on a 3-axis rubric and culled to 3-4. The output is presentation-ready: a shortlist with rationales, and an appendix of considered-and-rejected concepts that demonstrates process rigor.
Who this is for
Marketers and creative leads who need to present multiple campaign directions, not just one. Indie founders launching campaigns without a full agency. Agency strategists doing the first pass before involving the larger team. Anyone who has noticed that their “brainstorm meetings” produce three rewrites of the safe idea.
When to reach for it
When you need to present multiple campaign directions, not just one. Before booking a brainstorm meeting (use AI for divergence; use the meeting for convergence). When a stakeholder asks “what else did you consider?” and you have no good answer. Quarterly when you’re planning the campaign calendar and need to fill multiple slots without recycling.
Before you start
- Have your brief: audience, product, goal, constraints (budget, channels, timing), prior campaigns you’re explicitly NOT repeating. The “what we’re avoiding” list matters more than people think.
- Define what success looks like with one metric. “Increase signups by 20%” — not “increase awareness.” Vague goals produce vague scoring.
- Pick a high-context model: Claude Sonnet 4.6+ or GPT-5.5. Smaller / faster models collapse to the safe-middle on ideation tasks.
Step by step
- Paste your brief to the AI. Use this structure: audience persona (2-3 lines), product (1 paragraph), goal + metric, hard constraints, prior campaign list (“do not repeat the angles in any of these”). Without this scaffolding, the output is generic.
- Generate 12 concepts with forced differentiation:
Generate 12 campaign concepts for the brief above. Each
must take a different stance from this list:
emotional, functional, aspirational, contrarian,
community-driven, data-driven, urgency-based, founder-story,
educational, satirical, partnership-led, and platform-native.
Do not produce two concepts on the same stance.- For each concept, generate the briefing layer:
For each of the 12 concepts, give me:
- A 5-word campaign headline
- A 1-line creative hook
- The lead channel (Instagram / YouTube / email / OOH / etc.)
- One specific execution idea (not abstract — a real piece of content)- Score on 3 axes. Prompt: “Score each concept 1-5 on: on-brief (does it serve the goal), differentiation (vs prior campaigns and competitors), executability (can we ship in the budget). Be honest — the point is to surface real winners, not flatter.”
- Cull to top 3-4 by score. For each, ask for a 1-paragraph rationale: why this beats the eight you rejected.
- Stakeholder review with the shortlist + the appendix of 8-9 considered concepts. Showing the rejected ones with their scores demonstrates rigor and pre-answers “what about ___.”
- Save the brief + prompts as a template. Next campaign, swap the brief paragraph.
First-run exercise
- Run the workflow on a campaign you already shipped. Compare the 12 AI concepts to what you actually ran.
- Did the AI surface angles you hadn’t considered? If yes, that’s the workflow’s value. If no, your prompt didn’t enforce enough differentiation — sharpen the stance list.
- Score the AI concepts AS IF you were briefing a new campaign. Were the top 3 actually stronger than what you shipped? If the answer is “yes for 1 of 3,” that’s still a win.
- For the real next campaign, run the workflow before any brainstorm meeting. The meeting becomes about combining and stretching, not generating from scratch.
Quality check
- Are the 12 concepts genuinely different stances, or 12 variants of “we’re great”? If you can’t summarize each in 5 words that wouldn’t fit another concept, the differentiation collapsed — re-prompt with stricter labels.
- Does each “execution idea” name a real piece of content (a video format, a landing page type, a specific influencer category)? “Engaging content” is not an execution idea.
- Are the scores honest, or is the AI giving everything 4-5? If everything scores 4+, ask: “Be brutal — which of these would I actually be embarrassed to pitch?”
- Did the differentiation hold up after scoring, or did the top 3 all collapse to the safe-middle? If yes, your rubric is biased toward executability — adjust weights.
How to reuse this workflow
- Save the 12-stance prompt as a template. The stance list is the workflow’s secret — don’t lose it.
- Maintain a “considered angles” log per campaign so you don’t recycle stances next time. After 4 campaigns, you’ll see your stance habits (always picking founder-story; never picking satirical).
- For very specific industries (B2B SaaS, DTC, B2G), build stance lists tailored to your category. Generic 12-stance lists work; tailored ones work better.
Recommended workflow
A real launch: brief paragraph at 9am → 12 stances generated by 9:10 → headline + hook + channel + execution per concept by 9:25 → 3-axis scoring + rationales by 9:40 → top 3 + appendix saved to deck by 10am. Stakeholder meeting at 2pm gets a much sharper set of options. Total ideation time: 60 minutes (vs the 4-hour brainstorm it replaces).
Common mistakes
- Asking for “10 campaign ideas” — you get 10 variants of the safe one. Always force stance differentiation.
- Skipping the constraints section — the AI generates concepts that violate budget or channel reality, then you waste time discarding them.
- Not scoring — without a rubric, every idea feels equally good and you default to the loudest advocate’s pick.
- Generating without the “not repeating” list — you’ll quietly recycle last quarter’s campaign.
- Treating the AI’s scores as final — they’re a first pass; your context (politics, brand fit) overrides.
- Showing only the top 3 to stakeholders — the appendix of considered concepts is what makes you look strategic.
FAQ
- Which AI?: Claude Sonnet 4.6+ or GPT-5.5. Both handle 12-concept divergence well. Smaller models compress to the safe middle.
- Can it replace brainstorm meetings?: Not entirely. AI for divergence, meeting for convergence and stretching. The 12 concepts go INTO the meeting; the meeting picks the final winner.
- What if my industry is so regulated that half the stances don’t apply?: Build a tailored stance list for your category. Pharma, finance, and B2G all benefit from stance lists that fit constraints.
- How do I get the brief right?: Spend 20 minutes on the brief, not 5. Garbage in, garbage out. Audience persona + goal + metric + constraints + prior campaigns = the minimum.
- What about visual concepts, not just copy?: Add a step 5: for each top 3, generate a visual mood board. See the AI ad creative tutorial for the visual brief format.
- Will the same stance always win?: No. Different audiences and goals favor different stances. The workflow surfaces what fits THIS brief — your scoring catches it.