Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when the user needs a repeatable execution playbook for a product or offer launch: T-30 to D+7 timeline, per-function checklists for content, design, ads, sales, and engineering, launch-day war room, escalation matrix, and roll-back criteria. Trigger on 'launch playbook', 'launch process', 'launch checklist', 'launch SOP', 'war room', 'roll back launch', or 'how do we not fail this launch'.
.claude/skills/minhnv0807-60-launch-playbook-global/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 60% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-21 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 62% | 0% |
> A playbook is not a plan. A plan is for one launch; a playbook is reusable and improves after every launch. Successful launches are not luck — they are the product of correct preparation and disciplined execution. Run 59-go-to-market-global first for the GTM plan; this playbook is the execution layer underneath it.
Read the GTM plan and the campaign brief if available. If missing, ask up to 4 questions:
58-positioning-global or the GTM plan)41-campaign-asset-list-global)| Function | Checklist | |----------|-----------| | Content | Captions, scripts, and creative approved by leader; hashtags and UTM links complete; schedule set or a person on duty to post; backup content if creator delivery slips | | Design | Correct size per platform; assets follow the timeline T-14 brief > T-7 concept > T-5 complex > T-3 medium > T-1 fast; quick-asset standby on launch day | | Ads (media buyer) | Pixel and tracking verified; campaigns built and Paused; budget matches plan; UTMs on every link; backup creative ready in case of policy rejection | | Sales / support | Inbox coverage by shift and timezone; reply scripts for common comments and DMs; lead handoff process to sales; complaint-handling script | | Engineering | Page loads fast, forms work, payment tested; on-call engineer on launch day; redirect plan if the page goes down |
The war room is one shared channel plus a clear on-duty roster. The leader coordinates and each function has a single point of contact. Data is checked hourly and decisions are made same-day. For distributed teams, publish the roster in a single reference timezone with local equivalents, and require written status updates at each checkpoint so no decision depends on someone being awake.
| Hour (launch timezone) | Activity | Owner | |------------------------|----------|-------| | T-1h | Leader runs the final check (page, tracking, schedule) | Leader | | 0:00 | Publish the official announcement | Writer | | +0:30 | Turn on BOFU ads (Meta, TikTok, Google) | Media buyer | | +1:00 | Email and SMS blast: "We're live" | Leader | | +2:00 | Data check 1: page views, form submits, ad delivery | Leader | | +4:00 | Second post (different angle — social proof or urgency) | Writer | | +7:00 | Mid-day review — adjust or hold | Leader | | +12:00 | Evening push at peak time: stories, reposts | Writer | | +15:00 | Day-1 wrap: numbers, anomalies, decisions for day 2 | Leader |
| Day | Activity | Owner | |-----|----------|-------| | D+1 | Follow-up email to non-buying leads — handle objections | Leader | | D+1 | Mark winning and losing creative from day-1 data | Media buyer + designer | | D+2 | Post social proof: first buyers, early results | Writer | | D+3 | Media buyer report > leader decides scale, hold, or cut | Leader | | D+5 | Urgency push: "X seats left" or "bonus ends date]" | Writer + leader | | D+7 | Last call: email, SMS, post > close cart | Leader + writer |
| Situation | Severity | Action | Handled by | |-----------|----------|--------|------------| | Tracking not firing | Urgent | Pause ads, fix immediately | Media buyer + engineering | | Landing page down | Urgent | Redirect or fix within 30 minutes | Engineering | | Payment failing | Urgent | Notify engineering and leader immediately | Engineering | | Ads rejected by platform policy | High | Activate backup creative, file appeal | Media buyer | | CPA above 3x target | High | Notify leader immediately, wait for decision | Media buyer | | Negative comments spreading | High | Notify leader, do NOT self-handle — follow 66-crisis-playbook-global | Writer + leader | | Claim challenged or complaint alleging deceptive advertising | High | Pause the asset, escalate to leader and legal | Leader |
Roll-back means pausing part or all of the launch to fix it. The leader decides.
| Situation | Threshold | Roll-back action | |-----------|-----------|------------------| | Tracking wrong or not firing | Any | Pause all ads until fixed — do not burn budget blind | | CPA above threshold | > 3x target after the first 24h | Pause BOFU ads, keep organic; review creative and audience before restarting | | Landing page CVR far below expectation | < 1/3 of expected after 48h | Keep TOFU, pause BOFU push; fix headline, form, and speed, then restart | | Offer displayed wrong (price or bonus) | Any | Remove the post, fix, repost — faster than apologizing on the broken post | | Reputation issue or complaints spreading | Spreading | Switch to 66-crisis-playbook-global, stop all sales posts until resolved |
Roll-back is not failure. Continuing while knowing you are wrong is.
63-campaign-retrospective-global) — mandatory, never skipped because the team is busyFile name: launch-playbook-[product-name]-[YYYYMMDD].md — contains the T-30 to D+7 timeline with owners and deadlines filled in, the 5 function checklists, the war room roster, the escalation matrix, the roll-back criteria, and a "lessons applied from last launch" section.
59-go-to-market-global: run first — the GTM plan is the strategic input, this playbook is the execution layer.02-campaign-brief-global: brief the team before entering the timeline.41-campaign-asset-list-global: enumerate the assets to produce at the T-30 to T-21 checkpoint.63-campaign-retrospective-global: run after every launch — lessons feed back into this playbook.66-crisis-playbook-global: activate when negative sentiment spreads or a reputation issue appears.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 45,394 | 40,239 | -11% | 1 | 1 | 0% | 7,876 | 8,755 | +11% | 0 | 0 | — |
case-02 | fail→fail | 51,478 | 40,359 | -22% | 1 | 1 | 0% | 8,286 | 9,556 | +15% | 0 | 0 | — |
case-03 | fail→pass | 41,578 | 29,643 | -29% | 1 | 1 | 0% | 7,218 | 7,759 | +7% | 0 | 0 | — |
case-04 | pass→fail | 36,459 | 33,604 | -8% | 1 | 1 | 0% | 6,322 | 8,619 | +36% | 0 | 0 | — |
case-05 | pass→pass | 23,288 | 25,008 | +7% | 1 | 1 | 0% | 4,032 | 6,905 | +71% | 0 | 0 | — |
case-06 | pass→pass | 14,908 | 16,731 | +12% | 1 | 1 | 0% | 2,411 | 5,311 | +120% | 0 | 0 | — |
case-07 | pass→pass | 19,135 | 14,363 | -25% | 1 | 1 | 0% | 2,953 | 4,935 | +67% | 0 | 0 | — |
case-08 | pass→pass | 11,763 | 8,191 | -30% | 1 | 1 | 0% | 1,864 | 4,114 | +121% | 0 | 0 | — |
case-09 | fail→pass | 16,821 | 8,390 | -50% | 1 | 1 | 0% | 2,579 | 4,124 | +60% | 0 | 0 | — |
case-10 | fail→pass | 14,225 | 6,898 | -52% | 1 | 1 | 0% | 2,355 | 3,974 | +69% | 0 | 0 | — |
case-11 | pass→pass | 12,739 | 10,736 | -16% | 1 | 1 | 0% | 2,088 | 4,454 | +113% | 0 | 0 | — |
case-21 | fail→pass | 11,668 | 2,142 | -82% | 1 | 1 | 0% | 1,836 | 3,112 | +69% | 0 | 0 | — |
case-12 | fail→pass | 14,453 | 5,367 | -63% | 1 | 1 | 0% | 2,254 | 3,648 | +62% | 0 | 0 | — |
case-13 | fail→pass | 13,815 | 5,016 | -64% | 1 | 1 | 0% | 2,370 | 3,568 | +51% | 0 | 0 | — |
case-14 | pass→pass | 16,868 | 9,849 | -42% | 1 | 1 | 0% | 2,793 | 4,384 | +57% | 0 | 0 | — |
case-15 | fail→pass | 11,828 | 7,938 | -33% | 1 | 1 | 0% | 1,934 | 4,064 | +110% | 0 | 0 | — |
case-16 | fail→pass | 13,074 | 9,410 | -28% | 1 | 1 | 0% | 1,978 | 3,949 | +100% | 0 | 0 | — |
case-17 | fail→pass | 14,543 | 3,855 | -73% | 1 | 1 | 0% | 2,604 | 3,453 | +33% | 0 | 0 | — |
case-18 | fail→pass | 16,697 | 3,024 | -82% | 1 | 1 | 0% | 663 | 3,286 | +396% | 0 | 0 | — |
case-19 | fail→pass | 11,556 | 12,397 | +7% | 1 | 1 | 0% | 2,018 | 4,840 | +140% | 0 | 0 | — |
case-20 | fail→pass | 13,733 | 9,657 | -30% | 1 | 1 | 0% | 2,225 | 4,298 | +93% | 0 | 0 | — |
case-22 | fail→pass | 14,863 | 11,292 | -24% | 1 | 1 | 0% | 2,630 | 4,740 | +80% | 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, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +55 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is 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.