Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when the user needs a go-to-market plan for a new product, offer, or course: beachhead segment selection, launch messaging, channels by phase across pre-launch, launch, and post-launch, milestones, and go/no-go criteria. Trigger on 'go-to-market', 'GTM plan', 'product launch plan', 'launch plan', 'launching a new product', 'beachhead segment', or 'GTM strategy'.
.claude/skills/minhnv0807-59-go-to-market-global/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 88% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 86% | 0% |
> A launch is not a one-day event. It is a three-phase process: prepare, launch, sustain. Failed launches usually are not caused by a weak product — they are caused by nobody knowing, nobody believing, and nobody buying at the right moment. Run 58-positioning-global and 31-offer-design-global first to lock the message and the offer; run 60-launch-playbook-global afterwards for the detailed execution checklist.
Read .agents/product-marketing-context-global.md, the positioning doc, and the offer doc if available. If missing, ask up to 4 questions:
Score 2-3 candidate segments (1-5 per question, 45 max):
| Criteria group | Scoring questions | |----------------|-------------------| | Market need | Is the pain clear and growing? Is search or discussion volume high? Is the market durable rather than a passing trend? | | Competition | Are there large incumbents? Have they saturated the channels? Is there a niche nobody addresses clearly? | | Reachability | Do you already have a fitting channel or list? Can you test it on a small budget? Can the buyer decide quickly? |
Decision: 30+ points means launch on this beachhead. 25-30 means reconsider or narrow further. Below 25 means revisit the product or the segment.
Signals of a high-value beachhead customer: high LTV, low churn, active referrals. Filter question: "If this product disappeared tomorrow, what would you do?" The more painful the answer, the better the segment.
Offer summary (from 31-offer-design-global):
| Item | Detail | |------|--------| | Product name | | | List price / early bird / bundle | | | Sales window and capacity limit | | | Guarantee | | | USP — one sentence of difference | |
Launch message hierarchy — 5 layers, used in this order inside every asset:
Required audience insight: persona, pain point, desire, primary objection, buying trigger. In US/EU markets the dominant objection is usually credibility and reviews rather than price — plan proof accordingly.
| Channel | Phase | Content | KPI | |---------|-------|---------|-----| | Meta (Facebook + Instagram) | Pre + Launch | Teaser > announce > urgency | Reach, leads | | TikTok | Pre + Launch | Hook video, testimonial, creator content | Views, CTR | | Email + SMS (Klaviyo, Mailchimp, Attentive) | Pre + Launch | Warm-up sequence > offer > last call | Open rate, CVR | | Community seeding (Reddit, Discord, Slack, niche forums) | Pre-launch | Problem-aware content, no offer reveal | Engagement, saved posts | | Google Search / Shopping | Launch | High-intent and brand keywords | CVR, ROAS | | Affiliates / partners / creators | Launch | Co-promotion with tracked links | Attributed revenue |
Selection rule: you need both a REACH channel (drives traffic) and a CONVERT channel (closes). One channel rarely does both. Prioritize where the beachhead already spends time and where you have a real advantage.
PRE-LAUNCH (T-4 to T-1 weeks):
| Week | Activity | Owner | Output | |------|----------|-------|--------| | T-4 | Offer finalized, landing page live | Leader + dev | Page live, tracking verified | | T-3 | Produce teaser content (3-5 pieces) | Content | Posts ready | | T-2 | Start seeding and warm-audience ads | Content + performance | Engagement, custom audiences | | T-1 | Email/SMS pre-launch sequence | Leader | Warm list | | T-3 days | Ads built and verified, tracking checked | Media buyer | Campaigns in Paused state | | T-1 day | Full funnel final check > go/no-go | Leader | GO decision |
LAUNCH WEEK:
| Day | Activity | Channel | Owner | |-----|----------|---------|-------| | D-day morning | Official announcement | All | Leader | | D-day | Turn on BOFU ads | Meta, TikTok, Google | Media buyer | | D+1 | Follow-up email to non-buying leads | Email/SMS | Leader | | D+2 | Social proof — first customer reviews | Social | Writer | | D+3 | Media buyer report > scale, hold, or cut | — | Leader | | D+5 | Urgency push — remaining capacity or bonus deadline | All | Leader | | D+7 | Close cart — final email and post | Email + social | Leader |
POST-LAUNCH (D+8 to D+14): compile results (revenue, leads, ROAS, CPA), collect customer feedback to improve the product and capture testimonials, run the team retrospective, and decide whether to convert the campaign to evergreen.
| Milestone | Timing | Pass criteria | |-----------|--------|---------------| | M1 — Offer + page ready | T-4 | Page live, tracking verified, offer locked (price, bonuses, deadline) | | M2 — Audience warm | T-1 | Warm base large enough: X] engagers / Y] email subscribers / custom audience >= Z] | | M3 — Go/no-go | T-1 day | Passes the checklist below | | M4 — Mid-launch check | D+3 | CPA within threshold, first orders landed | | M5 — Close + retro | D+7 to D+14 | >= X]% of target hit, retrospective complete |
Go/no-go checklist (T-1 day, leader signs off):
Any missing item means NO-GO. Do not launch "roughly ready" and fix later.
| Metric | Target | How measured | |--------|--------|--------------| | Launch-week revenue | | | | Orders / seats sold | | | | CPL / CPA | | | | ROAS | | | | Email open rate | | | | Landing page CVR | | |
Derive targets backwards from the revenue goal using 10-reverse-kpi-global — never set targets by feel. Benchmark against references/benchmarks-global.md for the relevant region and industry; a US launch and an SEA launch are not comparable at face value.
File name: gtm-plan-[product-name]-[YYYYMMDD].md
markdown# GTM Plan — [Product] — Launch [date] ## 1. Product and offer summary | Name / price / early bird / capacity / guarantee / USP | ## 2. Beachhead segment - SOM score: [X]/45 — decision: - Persona + pain + desire + objection + trigger: ## 3. Positioning and messaging - Core message (one sentence): - Message hierarchy: Hook > Solution > Proof > Offer > Urgency ## 4. Channel plan by phase | Channel | Phase | Content | KPI | ## 5. Timeline [Pre-launch T-4 to T-1 | Launch week D-day to D+7 | Post-launch D+8 to D+14] | Checkpoint | Activity | Owner | Output | Deadline | ## 6. Milestones and go/no-go | Milestone | Timing | Criteria | - Go/no-go checklist T-1: ## 7. Success metrics | Metric | Target | How measured | ## 8. Plan B | Failure scenario | Fallback |
58-positioning-global: run first — the launch core message comes from the positioning statement.31-offer-design-global: run first — the offer must be locked before GTM planning.10-reverse-kpi-global: derive lead and budget targets from the revenue goal.61-budget-planning-global: allocate the launch budget across phases.60-launch-playbook-global: run after — day-by-day execution, war room, escalation.02-campaign-brief-global: turn the GTM plan into briefs for content and performance.63-campaign-retrospective-global)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 50,810 | 40,570 | -20% | 1 | 1 | 0% | 8,336 | 8,927 | +7% | 0 | 0 | — |
case-02 | fail→pass | 50,547 | 53,357 | +6% | 1 | 1 | 0% | 8,289 | 10,868 | +31% | 0 | 0 | — |
case-03 | fail→pass | 46,813 | 37,994 | -19% | 1 | 1 | 0% | 8,313 | 9,405 | +13% | 0 | 0 | — |
case-04 | pass→pass | 12,548 | 6,382 | -49% | 1 | 1 | 0% | 2,206 | 3,761 | +70% | 0 | 0 | — |
case-05 | fail→pass | 13,304 | 11,162 | -16% | 1 | 1 | 0% | 2,275 | 4,266 | +88% | 0 | 0 | — |
case-06 | pass→pass | 15,308 | 15,295 | -0% | 1 | 1 | 0% | 2,576 | 5,087 | +97% | 0 | 0 | — |
case-07 | pass→pass | 14,630 | 11,833 | -19% | 1 | 1 | 0% | 2,331 | 4,628 | +99% | 0 | 0 | — |
case-08 | pass→pass | 16,241 | 14,960 | -8% | 1 | 1 | 0% | 2,620 | 5,038 | +92% | 0 | 0 | — |
case-09 | fail→pass | 19,402 | 20,050 | +3% | 1 | 1 | 0% | 3,153 | 5,858 | +86% | 0 | 0 | — |
case-10 | fail→pass | 12,348 | 7,099 | -43% | 1 | 1 | 0% | 1,892 | 3,683 | +95% | 0 | 0 | — |
case-11 | pass→pass | 15,403 | 15,407 | +0% | 1 | 1 | 0% | 2,706 | 5,127 | +89% | 0 | 0 | — |
case-12 | fail→pass | 13,444 | 14,523 | +8% | 1 | 1 | 0% | 2,554 | 5,091 | +99% | 0 | 0 | — |
case-13 | pass→pass | 14,998 | 13,084 | -13% | 1 | 1 | 0% | 2,388 | 4,683 | +96% | 0 | 0 | — |
case-14 | pass→pass | 15,221 | 11,761 | -23% | 1 | 1 | 0% | 2,502 | 4,425 | +77% | 0 | 0 | — |
case-15 | pass→pass | 2,917 | 31,319 | +974% | 1 | 1 | 0% | 520 | 8,252 | +1487% | 0 | 0 | — |
case-16 | fail→fail | 14,489 | 11,593 | -20% | 1 | 1 | 0% | 2,364 | 4,334 | +83% | 0 | 0 | — |
case-17 | fail→pass | 13,384 | 10,201 | -24% | 1 | 1 | 0% | 2,195 | 4,315 | +97% | 0 | 0 | — |
case-18 | fail→pass | 14,371 | 12,489 | -13% | 1 | 1 | 0% | 2,223 | 4,543 | +104% | 0 | 0 | — |
case-19 | fail→pass | 8,874 | 2,650 | -70% | 1 | 1 | 0% | 1,492 | 3,020 | +102% | 0 | 0 | — |
case-20 | pass→pass | 20,893 | 27,671 | +32% | 1 | 1 | 0% | 3,688 | 7,322 | +99% | 0 | 0 | — |
case-21 | pass→fail | 30,026 | 40,990 | +37% | 1 | 1 | 0% | 4,344 | 9,851 | +127% | 0 | 0 | — |
case-22 | pass→fail | 27,451 | 32,334 | +18% | 1 | 1 | 0% | 4,742 | 7,914 | +67% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.