--- name: marketing-plan description: Build a 12-month marketing plan structured by AARRR, customized to budget, team, and funding stage. Use when the user says "marketing plan", "growth plan", "GTM plan", "90-day marketing roadmap", or "fractional CMO plan". Not for positioning and ICP capture: run product-marketing first. Not for single-channel execution: use the channel skill (emails, programmatic-seo, pricing, referrals). license: MIT metadata: author: TechTide AI (Alex Cinovoj) provenance: rewritten from patterns in an open SaaS marketing skills pack (MIT) category: Marketing & Growth --- # Marketing Plan A plan is only useful if the team can execute it with their actual budget, headcount, and stack. Generic plans fail. Every recommendation gets an AARRR stage tag, an owner, and the tooling that executes it. ## Workflow 1. **Intake.** Read `.agents/product-marketing.md` if it exists (run product-marketing first if not). Then gather: funding state and runway, team and who owns what, monthly budget by line, active channels and their status, past work to acknowledge, in-flight work and blockers, tool stack, unit economics (ARPC, retention, CAC). Full question list in references/intake.md. 2. **Score current state.** Rate each marketing surface 0-5 from available materials. Mark "scored from materials" so the team can push back with better data. 3. **Draft section by section, interactively.** Present each section, get approve/adjust/expand, save the confirmed version before moving on. Resumable: keep a progress file per client. Section structure in references/plan-template.md. 4. **Compile and verify.** Assemble the final markdown document. Check every cross-reference resolves, no machine paths ship, and voice matches the strategic frame. ## Structure (read references/plan-template.md before drafting) Thirteen sections: executive summary, strategic frame, current state, the five AARRR stages, 90-day roadmap, 12-month outlook, operations stack, tactical idea bank, measurement and open decisions. Brand and content are cross-cutting, not their own stage. AARRR in one line each: - Acquisition: strangers become aware - Activation: first valued experience (signup, onboarding, first session) - Retention: repeat use, churn prevention - Referral: retained users bring more users - Revenue: pricing, packaging, expansion ## Hard rules - **Budget honesty.** Anchor spend to funding stage. Do not plan paid spend a bootstrapped team cannot fund. Include explicit "what unlocks when the round closes" reasoning. - **Budget math, not vibes.** Revenue-based (5-40% of ARR) when CAC history exists, goal-based (reverse-engineer from the revenue target) when the goal is fixed. Add 10-20% experimental budget on top. - **CAC unknown is the highest-impact open decision.** Never gloss it. Every revenue projection depends on it. Blended CAC includes salaries, tools, and retainers, not just ad spend. - **Real growth is S-curves with plateaus**, not a hockey stick. Name the current phase, sequence the next curve while the current one still grows. - **First hire is a strategist, not a tactician.** Title conservatively (Lead, not VP). Outsource execution to contractors pre-Series-A. - **90-day roadmap has owners**, not just actions. - **Skipped ideas get a rationale.** What you are not doing, and why, is part of the plan. ## Verification Hand the executive summary to someone with no context. Expect them to state the 3 big bets, the 90-day priority, and the 12-month outcome in under 60 seconds of reading. If they cannot, the summary carries too little or too much, rewrite it. Then grep the final document for moves without an AARRR tag or an owner. Expect zero. Any found: tag or cut. ## Good vs Bad **Bad:** "Q2: improve SEO and build a community. Q3: scale paid." No owner, no budget check, no specific move, paid assumed before the round closes. **Good:** "Q2 (pre-raise, $1.5K/mo): ship 8 comparison pages targeting switchers (Acquisition, owner: founder, executes via programmatic-seo). Paid tests start only after seed closes, budgeted at $8K/mo (unlock noted in Section 10)." ## Completion checklist - [ ] Product marketing context read or created first - [ ] All 13 sections drafted and user-approved - [ ] Every move tagged with AARRR stage, owner, and executing skill/tool - [ ] Budget tied to funding stage with unlock reasoning - [ ] Open decisions listed explicitly, CAC status named - [ ] Past work acknowledged, skipped ideas have rationale - [ ] Exec summary passes the 60-second test Any box unchecked: not done. Fix or say so. ## Footguns - **Tactics without sequencing.** A list of 40 ideas is not a plan. Fix: 90-day roadmap in phases (unblock, foundation, velocity, compound), everything else parked in the idea bank with a status. - **Recommending beyond team capacity.** A 2-person team cannot run 6 channels. Fix: count hours per move, cut until it fits. - **Ignoring work already done.** Plans that skip past launches and content the team is proud of lose the room. Fix: acknowledge it in current state. - **Padding.** Dense beats long. If a section has nothing to say, write "not in scope for this 12-month plan" and move on.