you are an experienced business presentation consultant who helps fortune 500 executives craft concise, audience-focused slide decks for global stakeholders. generate a 12-slide ppt outline: title slide → agenda → 3–5 core sections → key takeaway → next steps. each slide has 1 title + 2 bullet points (≤12 words each, verb-started, quantified/deadline-driven). no full sentences. slide 4 compares q2 vs q3 metrics using % change only; slide 7 includes a 9-word risk mitigation callout. audience: non-technical regional managers in germany and japan — avoid idioms, unspelled acronyms, passive voice; cite “2024 q2 sales dashboard” for all data. replace “leverage”/“utilize” with “use”. after output, check: all titles start with action verbs; all bullets contain quantifiable results or deadlines — revise if not. then convert final outline into live-speaker notes (one paragraph per slide, no bullets/markdown).
☞☞☞AI 智能聊天, 问答助手, AI 智能搜索, 多模态理解力帮你轻松跨越从0到1的创作门槛☜☜☜
你需要用ChatGPT快速生成结构清晰、逻辑连贯、符合国际表达习惯的英文PPT大纲，而不是零散要点或中式英语堆砌。
AI一键生成成品PPT☜☜☜☜☜点击生成；
明确角色与场景
在提示词开头直接设定ChatGPT的身份：You are an experienced business presentation consultant who helps Fortune 500 executives craft concise, audience-focused slide decks for global stakeholders.
这句强制模型切换语境——它不再当通用AI，而是按专业顾问的思维组织信息，自动过滤冗余细节、预判听众关注点、规避文化盲区。
限定结构与粒度
要求输出必须包含Title Slide → Agenda → 3–5 core content sections → Key Takeaway → Next Steps，每页只写1个主标题+2个bullet points（每个point≤12词），禁用完整句子。
不写“Introduction”这种空泛页名，改用具体动作型标题，比如“Why This Matters to APAC Operations”；bullet points必须含动词开头，如“Cut onboarding time by 40% in Q3”而非“We have reduced onboarding time”。
注入真实约束条件
方法一：嵌入硬性参数
Include exactly 12 slides total. Slide 4 must compare Q2 vs Q3 metrics using % change only. Slide 7 requires a callout box with risk mitigation tactic — no more than 9 words.
方法二：绑定业务语境
Audience: non-technical regional managers in Germany and Japan. Avoid idioms, acronyms without first-use spelling, and passive voice. All data references must cite internal source “2024 Q2 Sales Dashboard”.
【不写“make it professional”这种无效指令】——模型无法解析抽象形容词，但能执行“replace ‘leverage’ with ‘use’”或“change all instances of ‘utilize’ to ‘use’”。
强制校验与迭代指令
第一步：生成初版大纲后，立即追加指令：
Check consistency: Does every section title start with an action verb? Does every bullet point contain at least one quantifiable result or deadline? If not, revise and re-output.
第二步：确认无误后，追加：
Now convert this outline into speaker notes — one short paragraph per slide, written as if speaking live to senior leaders, no markdown, no bullet points.
这一步让模型自我审查逻辑断层。很多用户跳过校验，直接拿初稿用，结果发现“Agenda”页列了5项，但正文只有3节——因为没触发验证机制。
