Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Playbook chuan hoa quy trinh launch san pham / offer: timeline T-30 → D+7, checklist tung bo phan (content, design, ads, sales, CSKH), war-room ngay launch, escalation matrix va tieu chi roll-back. Dung khi nhac 'launch playbook', 'quy trinh launch', 'checklist launch', 'war room'.
.claude/skills/minhnv0807-60-launch-playbook/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 68% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 33% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 120% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 107% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 189% | 0% |
> Playbook khac voi plan — playbook la quy trinh tai su dung duoc, cai tien sau moi lan launch. Launch thanh cong khong phai may man — la ket qua cua chuan bi dung va execution ky luat. Goi 59-go-to-market truoc de co GTM plan; playbook nay la lop thuc thi chi tiet.
Doc GTM plan va campaign brief neu co. Neu thieu, hoi toi da 4 cau:
58-positioning / GTM plan)41-campaign-asset-list)| Bo phan | Checklist | |---------|-----------| | Content | Caption/script/creative da duoc Leader approve; hashtag + link UTM day du; schedule da set hoac co nguoi truc dang; content backup neu UGC giao tre | | Design | Hinh dung size tung platform; asset theo timeline T-14 brief → T-7 concept → T-5 phuc tap → T-3 trung binh → T-1 nhanh; standby quick asset ngay launch | | Ads (Media Buyer) | Pixel/tracking verify xanh; ads setup dang Paused; budget dung plan; UTM day du tren moi link; backup creative san (phong ads reject) | | Sales / CSKH | Nguoi truc inbox theo ca; kich ban tra loi comment/inbox thuong gap; quy trinh chuyen lead cho sales; script xu ly complaint | | Dev / Ky thuat | LP load nhanh, form hoat dong, payment test; nguoi truc ky thuat ngay launch; phuong an redirect neu LP down |
War-room = 1 kenh chat chung + lich truc ro rang. Leader dieu phoi, moi bo phan co 1 dau moi. Data check theo gio, quyet dinh trong ngay.
| Gio | Activity | Owner | |-----|----------|-------| | 7:00 | Leader check lai toan bo lan cuoi (LP, tracking, lich dang) | Leader | | 8:00 | Dang announce post chinh thuc | Writer | | 8:30 | Bat BOFU ads (Meta, TikTok) | Media Buyer | | 9:00 | Zalo OA / Email blast: "Mo roi!" | Leader | | 10:00 | Check data dot 1: LP views, form submits, ads delivery | Leader | | 12:00 | Post lan 2 (angle khac — social proof / urgency) | Writer | | 15:00 | Review so lieu giua ngay — can adjust khong | Leader | | 20:00 | Evening push (peak time): story, repost | Writer | | 23:00 | Tong ket ngay 1: so lieu, anomalies, quyet dinh ngay 2 | Leader |
| Ngay | Activity | Owner | |------|----------|-------| | D+1 | Email follow-up leads chua mua — xu ly objection | Leader | | D+1 | Danh dau creative winner/loser tu data ngay dau | Media Buyer + Designer | | D+2 | Post social proof: first buyers / early results | Writer | | D+3 | Media buyer report → Leader quyet: tang/giu/scale | Leader | | D+5 | Urgency push: "Con X slot / het bonus ngay Y" | Writer + Leader | | D+7 | Last call: email + post + Zalo → close cart | Leader + Writer |
| Tinh huong | Muc do | Action | Nguoi xu ly | |-----------|--------|--------|-------------| | Tracking khong fire | Khan | Dung ads, fix ngay | Media Buyer + Dev | | LP down | Khan | Redirect hoac fix trong 30 phut | Dev | | Payment loi | Khan | Bao Dev + Leader ngay | Dev | | Ads bi reject | Cao | Bat backup creative | Media Buyer | | CPL > 3x target | Cao | Bao Leader ngay, cho quyet dinh | Media Buyer | | Comment tieu cuc viral | Cao | Bao Leader, KHONG tu xu ly — theo 66-crisis-playbook | Writer + Leader |
Roll-back = tam dung 1 phan hoac toan bo launch de sua. Leader la nguoi quyet.
| Tinh huong | Nguong | Hanh dong roll-back | |-----------|--------|---------------------| | Tracking sai / khong fire | Bat ky | Pause toan bo ads den khi fix xong — khong dot budget mu | | CPL / CPMess vuot nguong | > 3x target sau 24h dau | Pause ads BOFU, giu organic; review creative + audience truoc khi bat lai | | LP CVR qua thap | < 1/3 muc ky vong sau 48h | Giu ads TOFU, tam dung push BOFU; fix LP (headline, form, toc do) roi bat lai | | Offer sai (gia/bonus hien thi sai) | Bat ky | Go post loi, sua, dang lai — uu tien sua nhanh hon xin loi tren post cu | | Su co uy tin / complaint lan rong | Dang lan | Chuyen sang 66-crisis-playbook, dung moi post ban hang den khi xu ly xong |
Roll-back KHONG phai that bai — chay tiep khi biet dang sai moi la that bai.
63-campaign-retrospective) — buoc bat buoc, khong bo qua vi banTen file: launch-playbook-[ten-san-pham]-[YYYYMMDD].md — gom: timeline T-30 → D+7 da dien owner + deadline, checklist 5 bo phan, war-room roster, escalation matrix, tieu chi roll-back, muc "bai hoc lan truoc da ap dung".
59-go-to-market: chay truoc — GTM plan la dau vao chien luoc, playbook la lop thuc thi.02-brief-chien-dich: brief campaign cho team truoc khi vao timeline.41-campaign-asset-list: liet ke asset can san xuat o moc T-30 → T-21.63-campaign-retrospective: chay sau moi launch — bai hoc update nguoc vao playbook.66-crisis-playbook: kich hoat khi comment tieu cuc viral hoac su co uy tin.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 30,776 | 38,701 | +26% | 1 | 1 | 0% | 5,215 | 8,738 | +68% | 0 | 0 | — |
case-02 | fail→pass | 46,823 | 47,147 | +1% | 1 | 1 | 0% | 8,290 | 11,001 | +33% | 0 | 0 | — |
case-03 | fail→fail | 39,184 | 36,819 | -6% | 1 | 1 | 0% | 6,386 | 8,812 | +38% | 0 | 0 | — |
case-04 | fail→fail | 21,860 | 33,184 | +52% | 1 | 1 | 0% | 3,618 | 8,204 | +127% | 0 | 0 | — |
case-05 | fail→pass | 12,631 | 12,687 | +0% | 1 | 1 | 0% | 2,051 | 4,516 | +120% | 0 | 0 | — |
case-06 | fail→pass | 14,727 | 11,366 | -23% | 1 | 1 | 0% | 2,233 | 4,620 | +107% | 0 | 0 | — |
case-07 | pass→pass | 13,202 | 11,948 | -9% | 1 | 1 | 0% | 2,189 | 4,665 | +113% | 0 | 0 | — |
case-08 | fail→pass | 10,118 | 11,601 | +15% | 1 | 1 | 0% | 1,651 | 4,773 | +189% | 0 | 0 | — |
case-09 | pass→pass | 15,271 | 12,693 | -17% | 1 | 1 | 0% | 2,342 | 4,911 | +110% | 0 | 0 | — |
case-10 | fail→pass | 13,638 | 9,699 | -29% | 1 | 1 | 0% | 2,231 | 4,212 | +89% | 0 | 0 | — |
case-11 | fail→pass | 17,110 | 24,071 | +41% | 1 | 1 | 0% | 2,908 | 7,045 | +142% | 0 | 0 | — |
case-12 | fail→fail | 15,763 | 19,162 | +22% | 1 | 1 | 0% | 2,541 | 5,861 | +131% | 0 | 0 | — |
case-13 | pass→pass | 12,973 | 11,484 | -11% | 1 | 1 | 0% | 2,214 | 4,696 | +112% | 0 | 0 | — |
case-14 | fail→pass | 17,501 | 21,138 | +21% | 1 | 1 | 0% | 2,906 | 6,269 | +116% | 0 | 0 | — |
case-15 | pass→pass | 9,728 | 5,557 | -43% | 1 | 1 | 0% | 1,470 | 3,657 | +149% | 0 | 0 | — |
case-16 | fail→pass | 12,253 | 15,631 | +28% | 1 | 1 | 0% | 1,943 | 5,210 | +168% | 0 | 0 | — |
case-17 | pass→pass | 13,395 | 12,972 | -3% | 1 | 1 | 0% | 2,000 | 4,749 | +137% | 0 | 0 | — |
case-18 | fail→pass | 13,258 | 14,145 | +7% | 1 | 1 | 0% | 2,042 | 4,921 | +141% | 0 | 0 | — |
case-19 | pass→fail | 25,476 | 33,183 | +30% | 1 | 1 | 0% | 4,071 | 8,229 | +102% | 0 | 0 | — |
case-20 | pass→pass | 22,195 | 24,674 | +11% | 1 | 1 | 0% | 3,444 | 6,664 | +93% | 0 | 0 | — |
case-21 | pass→pass | 16,898 | 18,002 | +7% | 1 | 1 | 0% | 2,489 | 5,376 | +116% | 0 | 0 | — |
case-22 | pass→pass | 13,412 | 11,453 | -15% | 1 | 1 | 0% | 1,988 | 4,563 | +130% | 0 | 0 | — |
case-23 | pass→fail | 13,529 | 19,017 | +41% | 1 | 1 | 0% | 2,266 | 5,945 | +162% | 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. 23 cases were attempted. The headline lift of +35 percentage points is the difference between those two pass rates over the 23 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.