Design Amazon keyword-ranking experiments that identify promising terms from relevance, conversion, current organic visibility, and placement performance, then sequence long-tail, mid-tail, and head-term tests. Use when the user asks which keywords to push, what an ad position may reveal about order potential, or why organic rank stalls despite paid orders. Do not equate sponsored placement with organic rank or guarantee movement.
npx skills add https://github.com/xjli360/sealeap-amazon-skills --skill sealeap-amazon-keyword-ranking-experiment
用广告实验筛选与商品最匹配、在合理位置能产生利润的关键词,再按证据逐级扩大,而不是把排名当作可直接购买的结果。
缺失的数据要明确列为缺口,并把结论标成事实、估算或假设;不要补造数字。
只有在本任务确实需要外部市场、竞品、关键词或公开网页证据时,才读取 references/mcp-data-plan.md,并使用 scripts/mcp_research.py。
从真实相关、已有转化、自然可见度较好或广告效率较高的词中建立候选集,不只按当前名次。
选择一个或多个广告位类别,固定商品页、价格和预算,明确不是精确页码控制。
比较各位置的曝光、CTR、CVR、CPA 和贡献利润,形成区间而非把广告订单直接当作未来自然订单。
优先给高意图长尾或中部词稳定预算;达到利润和样本门槛后才扩大。
检查相对转化、点击、库存、价格、竞争和词根相关性;不要仅靠更高竞价追自然位。
当多个相关词形成稳定基本盘后,小规模测试核心词,并保留止损和退出路径。
结尾列出数据窗口、关键假设、证据缺口、风险和下一步;如包含待执行动作,单独放在“待批准变更”中。
Take xjli360/sealeap-amazon-keyword-ranking-experiment from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.