Steps to build your own AI strategist
This resource is based on She taught AI her process. Then it scaled her business., featuring Barbara Carneiro of Spice Ink, published on the AI Lab by ActiveCampaign.
Get the checklist
By the end of this checklist, you’ll have an AI strategist that thinks like you, draws from your institutional knowledge, and produces strategy roadmaps in minutes instead of hours. It’s based on the system that Barbara Carneiro built at Spice Ink—expect 4–6 hours of initial setup, with ongoing refinement as you feed it more context.
Before you start
- Set up a ChatGPT Plus account: you’ll need access to custom GPTs (chatgpt.com)
- Set up a Dropbox account with Dash access: this will be your knowledge base (dash.dropbox.com)
- Gather your existing strategy materials: discovery notes, brand boards, strategy decks, client folders, meeting recordings
The workflow
Phase 1: Define your reasoning framework
After this phase, you’ll have: a documented strategic framework that captures how you make decisions
- Identify the patterns in your strategic thinking: write down the 3–5 dimensions you evaluate when making brand or marketing decisions. Carneiro’s ARC Method uses Archetype (brand personality), Reality (audience emotional landscape), and Chasm (the communication gap between them).
- Define the options within each dimension: list every variation. Carneiro maps 12 archetypes, 4 active generations, and 9 communication gaps—producing 432 possible strategic combinations.
- Write descriptions for each option: include behavioral patterns, examples, and distinguishing characteristics so the AI can differentiate between them.
- Document your decision rules: explain how you select from the combinations. What signals tell you to pick one archetype over another?
“I didn’t want it to create content. I wanted it to think like me.”
Phase 2: Build the AI engine
After this phase, you’ll have: a custom GPT that understands your framework and can apply it to client work
- Create a custom GPT in ChatGPT: go to “Explore GPTs” and select “Create.” Name it something meaningful—Carneiro named hers “Clarity.”
- Upload your framework as system instructions: paste your reasoning framework (from Phase 1) into the custom GPT’s configuration. Include the dimensions, options, decision rules, and example outputs.
- Test with a real scenario: give the GPT a client brief and ask it to produce a strategy recommendation using your framework.
- Refine the output: compare the GPT’s recommendation to what you would have produced. Adjust the system instructions where the GPT missed your reasoning.
- Repeat with 3–5 different client types: the more varied your test cases, the better the GPT learns your patterns.
Framework prompt
Using the [FRAMEWORK_NAME], analyze [CLIENT_NAME]‘s brand. Their brand personality aligns with [ARCHETYPE]. Their target audience is [GENERATION/SEGMENT]. Identify the communication gap and recommend a strategy direction.
Phase 3: Connect the knowledge base
After this phase, you’ll have: an AI strategist that draws from your full institutional memory
- Organize your Dropbox by client: create a folder per client containing discovery notes, strategy decks, brand boards, newsletters, and meeting recordings.
- Enable Dash search across your workspace: Dash indexes your files, including auto-transcribed video that is searchable by keyword and timestamp.
- Link Dash to your custom GPT workflow: use Dash to retrieve client-specific context, then feed that context to your GPT for strategy generation.
- Test the full pipeline: run a complete strategy roadmap from Dash context retrieval through GPT output. Compare against a roadmap you produced manually.
- Set up team access: grant your team access to Dash so they can query the knowledge base directly, reducing your role as the bottleneck.
Strategy generation
Here is the context for [CLIENT_NAME] from our knowledge base: [PASTE DASH SEARCH RESULTS]. Based on this context and the [FRAMEWORK_NAME], produce a strategy roadmap that includes brand positioning, audience targeting, and communication recommendations.
Quick reference
- Total time: 4–6 hours for initial setup, ongoing refinement
- Tools needed: ChatGPT (Plus plan), Dropbox Dash
- Key output: an AI strategist that produces strategy roadmaps in ~15 minutes instead of ~6 hours
Ready for the full story?
Read She taught AI her process. Then it scaled her business. featuring Barbara Carneiro of Spice Ink, published on the AI Lab by ActiveCampaign.
Related
More data from the AI Lab.