← ClaudeAtlas

marketing-councillisted

Evaluate a marketing plan, campaign brief, GTM strategy, launch plan, positioning doc, or funnel through a council of 6 specialist advisors arranged in three opposing pairs (Brand Strategist vs Performance Marketer, Customer vs Rival, Creative Director vs Operator). Each scores the plan independently, then they peer-review each other anonymously, then a CMO chairman issues a Go / Go-with-changes / Rework / Kill verdict with a health scorecard. MANDATORY TRIGGERS: 'marketing council', 'run the marketing council', 'CMO review', 'review my marketing plan', 'evaluate this campaign', 'pressure-test this campaign', 'red team this launch', 'score this plan'. STRONG TRIGGERS (use when the user supplies an actual plan, brief, or strategy document): 'what do you think of this marketing plan', 'is this campaign any good', 'will this launch work', 'critique my GTM', 'evaluate my positioning', 'should we run this campaign', 'poke holes in this'. Do NOT trigger for writing marketing copy, drafting a single ad, answering ge
zihadizisan/marketing-council · ★ 0 · AI & Automation · score 67
Install: claude install-skill zihadizisan/marketing-council
# Marketing Council A marketing plan reviewed by one AI gets one opinion, and that opinion bends to how the plan was framed. Present a campaign enthusiastically and you get validation. Present the same campaign anxiously and you get concerns. Same plan, opposite readings. The Marketing Council fixes this by running the plan past six advisors who are structurally incapable of agreeing with each other, making them review each other's work blind, and then having a chairman force the disagreement into a single verdict. This is the LLM Council methodology (Karpathy) retargeted for marketing evaluation. The generic council uses five thinking styles for open decisions. This one uses six marketing lenses, adds a structured intake gate, forces numeric scoring, and ends in a four-way verdict instead of a recommendation. --- ## The design principle The roster is built on **tensions, not job titles.** The failure mode when building a marketing council is staffing it by channel — a Social person, an SEO person, an Email person, a Content person. That produces six advisors who each say "yes, and here's how my channel contributes." No tension, no signal. These six are arranged in three opposing pairs. Each pair argues about something marketers genuinely and legitimately disagree about. If two advisors start producing similar output, that is a roster failure to correct, not a consensus to celebrate. --- ## The roster ### Pair 1 — Time horizon: equity vs. revenue **1. The Brand St