Prompts to train an AI strategist
This resource is based on She taught AI her process. It scaled her business., featuring Barbara Carneiro of Spice Ink, published on the AI Lab by ActiveCampaign.
Get the prompt template
How to use these prompts
These are ready-to-use prompts based on the workflow that Barbara Carneiro built at Spice Ink. Copy, paste, and swap in your details where you see [BRACKETS]. Every AI tool and model behaves a little differently, so treat what comes back as a starting point—review the output and refine from there.
Prompt 1: Extract your strategic framework
- Best for: Documenting the decision patterns you apply intuitively so an AI can replicate them
- Use with: ChatGPT (custom GPT configuration) or Claude
Extract your strategic framework
I’m a [YOUR_ROLE] at [YOUR_COMPANY/AGENCY]. When I evaluate a brand’s communication strategy, I consider these dimensions:
- Brand personality: [LIST YOUR DIMENSIONS—e.g., archetypes, brand values, voice characteristics]
- Audience profile: [LIST YOUR AUDIENCE SEGMENTS—e.g., generational, behavioral, psychographic]
- Communication gap: [DESCRIBE THE GAP YOU IDENTIFY—e.g., where the message breaks down between brand intent and audience perception]
For each dimension, here are the options I work with:
Brand personality options: [LIST ALL OPTIONS WITH SHORT DESCRIPTIONS]
Audience segments: [LIST ALL SEGMENTS WITH BEHAVIORAL PATTERNS]
Communication gaps: [LIST ALL GAP TYPES WITH EXAMPLES]
Your job is to act as my strategic thinking partner. When I give you a client brief, analyze it through these three dimensions and recommend a strategy direction. Always explain your reasoning by mapping back to the specific options that you selected and why.
Variables to fill in:
- [YOUR_ROLE]—your title (e.g., “brand strategist,” “agency founder,” “marketing director”)
- [YOUR_COMPANY/AGENCY]—your company name
- [LIST YOUR DIMENSIONS]—the strategic dimensions you evaluate (Carneiro uses Archetype, Reality, Chasm)
- [LIST ALL OPTIONS]—every option within each dimension, with descriptions
What to expect: The AI should confirm that it understands your framework and ask clarifying questions if any dimension is vague. Test by giving it a client brief immediately after.
Prompt 2: Generate a strategy roadmap from a client brief
Best for: Producing a strategy recommendation after a discovery session
Use with: ChatGPT (custom GPT with framework loaded) or Claude (with framework pasted in)
Strategy roadmap
Here is the brief for [CLIENT_NAME]:
Company: [COMPANY_DESCRIPTION]
Industry: [INDUSTRY]
Target audience: [PRIMARY_AUDIENCE_DESCRIPTION]
Business goals: [WHAT_THEY_WANT_TO_ACHIEVE]
Current brand positioning: [HOW_THEY_CURRENTLY_PRESENT_THEMSELVES]
Context from our knowledge base:
[PASTE RELEVANT NOTES FROM DISCOVERY SESSIONS, PAST WORK, OR DROPBOX DASH SEARCH RESULTS]
Using my strategic framework, produce a strategy roadmap that includes:
- Recommended brand personality classification and reasoning
- Audience segment analysis with behavioral patterns
- Identified communication gap and how to bridge it
- Three specific tactical recommendations based on this strategy
Variables to fill in:
- [CLIENT_NAME]—the client you’re strategizing for
- [COMPANY_DESCRIPTION]—1–2 sentences about what they do
- [PRIMARY_AUDIENCE_DESCRIPTION]—who they serve
- [PASTE RELEVANT NOTES]—pull from your knowledge base (Dropbox Dash, past strategy decks, discovery session transcripts)
What to expect: A structured strategy recommendation that maps to your framework. Carneiro’s system produces these in about 15 minutes—down from the 6 hours she used to spend doing it manually.
Follow-up prompt
Compare this strategy recommendation to [COMPETITOR_OR_PREVIOUS_CLIENT] in the same industry. Where does this client’s communication gap differ, and how should the tactical recommendations shift to account for that?
Prompt 3: Simulate audience experience
Best for: Testing whether a deliverable (email, landing page, ad) matches the target audience’s perception
Use with: ChatGPT (with image analysis enabled) or Claude
Simulate audience experience
Please apply a [TARGET_AUDIENCE] readability simulation to this [ASSET_TYPE]. The goal is to simulate how [TARGET_AUDIENCE_DESCRIPTION] might experience this content.
Consider:
- Visual accessibility (font size, contrast, layout complexity)
- Language complexity (jargon, sentence length, assumed knowledge)
- Emotional resonance (does the tone match what this audience responds to?)
- Communication gap (based on the [ARCHETYPE] brand personality targeting [AUDIENCE_SEGMENT])
Provide:
- A description of what the target audience would actually experience
- Specific elements that work well for this audience
- Specific elements that create friction
- Recommended changes ranked by impact
Variables to fill in:
- [TARGET_AUDIENCE]—the specific audience segment (e.g., “older user,” “Gen Z buyer,” “C‑suite executive”)
- [TARGET_AUDIENCE_DESCRIPTION]—behavioral and perceptual traits of this segment
- [ASSET_TYPE]—what you’re testing (e.g., “email design,” “landing page,” “social ad”)
- [ARCHETYPE]—the brand personality from your framework
- [AUDIENCE_SEGMENT]—the audience classification from your framework
What to expect: A detailed analysis of how the content lands for that audience, with specific friction points and fixes. Carneiro uses this to test client deliverables: in one example, she simulated how an elderly user would experience a website with thin fonts and low contrast.
Prompt 4: Train your team to query the AI strategist
Best for: Giving team members a standard format for getting strategy answers without going through you
Use with: Any LLM (ChatGPT, Claude, Gemini)
Query your AI strategist
I need a strategy check on [DELIVERABLE_TYPE] for [CLIENT_NAME].
Client context: [1–2 SENTENCES ABOUT THE CLIENT AND THEIR STRATEGY]
The deliverable: [DESCRIBE WHAT YOU’RE REVIEWING]
My concern: [SPECIFIC QUESTION OR UNCERTAINTY]
Check this against the client’s strategy profile and tell me:
- Whether it aligns with the brand personality framework
- Whether it speaks to the identified audience segment
- Whether it addresses or widens the communication gap
- What I should change before this goes live
Variables to fill in:
- [DELIVERABLE_TYPE]—newsletter, social post, ad copy, landing page, etc.
- [CLIENT_NAME]—the client
- [1–2 SENTENCES]—quick context the AI needs (or reference the knowledge base)
- [SPECIFIC QUESTION]—what’s bugging you about this piece
What to expect: A focused alignment check that tells the team member whether to proceed, revise, or escalate. This is the prompt that replaces the bottleneck—instead of waiting for the agency owner to review, the team queries the AI.
Tips for better results
- Start by loading your full framework into a custom GPT’s system instructions—this saves you from pasting it every time
- Connect your knowledge base (Dropbox Dash or similar) so the AI draws from real client history, not generic advice
- Run Prompt 2 on three past clients first and compare the output to what you actually delivered—this reveals where your framework documentation has gaps
- Give the AI access to meeting transcripts and discovery notes, not polished decks—raw context produces better strategic reasoning
Ready for the full story?
Read She taught AI her process. It scaled her business., featuring Barbara Carneiro of Spice Ink, published on the AI Lab by ActiveCampaign.
Related
More data from the AI Lab.