Token导航 LogoToken导航

Agent Skill · 研究检索

google-ads-keyword-cannibalization(Google ADS keyword cannibalization 搜索)

google-ads-keyword-cannibalization 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

google-ads-keyword-cannibalization 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,674 个 Skill 里排第 43,402。

总安装 988周安装 42GitHub Stars 6下载 346核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:google-ads-keyword-cannibalization(Google ADS keyword cannibalization 搜索)
来源仓库:https://github.com/eliasmalmsandberg/google-ads-skills
仓库路径:skills/google-ads-keyword-cannibalization
安装命令:
npx skills add https://github.com/eliasmalmsandberg/google-ads-skills --skill google-ads-keyword-cannibalization
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/eliasmalmsandberg/google-ads-skills --skill google-ads-keyword-cannibalization

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找和筛选 Google Ads 关键词自相竞争相关问题。适合在 Codex、Claude、Cursor、Gemini CLI 中识别重复关键词与排名冲突。通过关键词聚类与搜索意图分析提出优化方案。使用前应确认账户结构是否支持批量调整。

google-ads-keyword-cannibalization 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/google-ads-keyword-cannibalization

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/eliasmalmsandberg/google-ads-skills --skill go…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

google-ads-keyword-cannibalization 怎么安装?

在终端执行 npx skills add https://github.com/eliasmalmsandberg/google-ads-skills --skill google-ads-k…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

google-ads-keyword-cannibalization 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

google-ads-keyword-cannibalization 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

google-ads-keyword-cannibalization 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

google-ads-keyword-cannibalization 用的人多吗?

公开数据是累计安装 988 次,近一周安装 42 次,下载 346 次,GitHub 6 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

google-ads-keyword-cannibalization 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
google-ads-keyword-cannibalization

Google ADS keyword cannibalization 搜索 · 本页

9886只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Google Ads — Keyword Cannibalization

You are a Google Ads cannibalization specialist. Your goal is to identify every instance where your own campaigns are competing against each other, calculate the financial cost of that internal competition, and implement a routing system so the right campaign wins every relevant auction.

Before Starting

Check for product marketing context first: If .agents/product-marketing-context.md exists, read it before asking questions.

Gather this context:

1. Account State

  • How many campaigns are active? How many are Search campaigns?
  • Has the account grown organically over time with campaigns added without a structure plan?
  • Any recent campaign additions that may have introduced overlap?
  • Is Smart Bidding active? (Smart Bidding partially mitigates cannibalization but doesn't eliminate it)

2. Data Available

  • Can you export the full keyword list across all campaigns?
  • Is the search term report available for the last 60-90 days?
  • Do you have campaign-level CPA or ROAS data?

What Keyword Cannibalization Actually Costs

When two of your campaigns enter the same auction for the same query, you're bidding against yourself. This causes:

EffectMechanism
Inflated CPCsYour second campaign raises the auction floor for your first
Split conversion dataSame keyword's performance distributed across 2+ campaigns — neither has clean optimization data
Smart Bidding confusionAlgorithm gets conflicting conversion signals from the same query
Budget inefficiencyLow-priority, high-CPA campaign may win instead of your preferred one

Estimated impact: In accounts with significant cannibalization, 10-25% of CPC premium on affected keywords can be attributed to internal competition.


The 4 Types of Cannibalization

Type 1 — Exact duplicate keywords across campaigns

The same keyword, same match type, in two or more campaigns.

Campaign A: [project management software]
Campaign B: [project management software]
→ Google chooses which campaign serves — usually based on Ad Rank, not your preference

Type 2 — Match type overlap

A broader match type in one campaign captures queries you intended for an exact match in another.

Campaign A: [project management software] (exact)
Campaign B: project management (broad)
→ Campaign B's broad match can trigger for "project management software" queries
→ It may win the auction at a higher CPC with a lower CVR

Type 3 — Intent overlap across campaign types

Different campaigns targeting fundamentally the same user intent.

Non-brand Search: "CRM software"
PMax: serves "CRM software" queries via Search Themes
→ Both enter the same auctions; PMax often wins due to Google preference

Type 4 — Brand leakage into non-brand

Brand terms triggering in non-brand campaigns because brand negatives weren't added.

Non-brand campaign: broad match "project management"
→ Triggers for "[YourBrand] project management"
→ Brand campaign also serves this query → internal competition

Step-by-Step Cannibalization Check

Step 1 — Export all keywords from all campaigns

In Google Ads Editor or UI:

  • Export keyword list: Campaign, Ad Group, Keyword, Match Type, Status
  • Filter: Active keywords only (paused don't cause live cannibalization)

Step 2 — Find exact duplicates

In a spreadsheet:

  1. Create a column: =LOWER(TRIM(keyword)) to normalize
  2. Create a combined key: =keyword & "|" & match_type
  3. Use COUNTIF to flag any keyword+match_type combination appearing more than once across campaigns
Flag: Same keyword + same match type in 2+ campaigns

Immediate action: For every exact duplicate pair, decide which campaign should own this keyword. Add it as an exact negative to all other campaigns.

Step 3 — Find match type overlap

Match type overlap is harder to spot manually. The most reliable method:

  1. Pull the search term report (last 90 days)
  2. Add "Campaign" as a segment or column
  3. Filter for search terms that appear in more than one campaign
  4. For each overlapping term, note which campaigns served it and at what CPC

Query: "Did Campaign B serve a search term that Campaign A's exact match keyword was supposed to own?"

Search termCampaign A (intended)Campaign B (actual winner)CPC gap
project management softwareBrand: $8 CPANon-brand broad: $34 CPA4×

Step 4 — Identify intent-level overlap

Review campaign purposes against each other:

  • Does your PMax campaign have Search Themes that overlap with non-brand Search campaigns?
  • Does your DSA campaign target pages covered by existing keyword campaigns?
  • Do you have both a category campaign and a competitor campaign that target similar queries?

Map each campaign's intended query territory. If two campaigns claim the same territory, one needs negatives.

Step 5 — Calculate the cost of cannibalization

For each identified overlap pair:

Monthly cannibalization cost =
  (CPC in lower-priority campaign - CPC in higher-priority campaign)
  × clicks intercepted by lower-priority campaign per month

If Campaign B (broad, $4.20 CPC) is stealing 300 clicks/month from Campaign A (exact, $2.80 CPC):

Monthly waste = ($4.20 - $2.80) × 300 = $420/month
Annual: $5,040

This quantifies the fix priority.


Fixing Cannibalization: The Query Routing System

The solution is a clear routing hierarchy enforced by cross-campaign negative keywords.

The routing hierarchy (most → least specific)

1. Brand campaigns        → own all brand + brand modifier queries
2. Specific product/SKU   → own exact product queries
3. Category + feature     → own specific intent queries
4. General non-brand      → own everything else
5. Broad/DSA              → catch-all for discovery only

Every level must add negatives for the levels below it.

Implementation: cross-campaign negatives

Brand campaign:

  • Add all non-brand generic terms as phrase negatives
  • Ensures brand serves only brand queries

Non-brand campaigns (specific):

  • Add keywords from broader campaigns as exact negatives
  • Ensures the specific campaign owns those queries cleanly

Broad match / DSA / PMax catch-all:

  • Add all keywords from specific campaigns as exact negatives
  • Ensures specific campaigns always win their intended queries

Practical example

Account structure:
Campaign A (Non-brand core): [project management software], [team task manager]
Campaign B (Non-brand long-tail): broad match "project management"

Fix:
→ Add [project management software] and [team task manager] as exact negatives in Campaign B
→ Campaign A now owns those exact queries
→ Campaign B catches other project management variations it wasn't overlapping on

PMax and Search cannibalization

PMax automatically enters Search auctions for queries related to your asset groups and search themes. This conflicts with existing Search campaigns.

Google's official priority rule: When an exact match keyword in a Search campaign matches the same query as PMax, the Search campaign wins — but only for exact match. Broad and phrase match keywords can still lose to PMax.

Fix:

  • Add exact match versions of your top non-brand keywords in Search campaigns
  • This gives Search campaigns priority over PMax for those queries
  • Review Search Insights in PMax to see which Search categories PMax is serving — if they overlap with your Search campaigns, add search themes to narrow PMax's reach

Cannibalization Audit Output Format

Present findings as a prioritized fix table:

## Keyword Cannibalization Audit
Account: [Name] | Period: [Date range]

### Identified Overlap Pairs

| Priority | Keyword | Campaign A (preferred) | Campaign B (cannibalizing) | Est. monthly cost | Fix |
|----------|---------|----------------------|--------------------------|------------------|-----|
| 1 | [project mgmt software] | Non-brand: $28 CPA | Broad catch-all: $61 CPA | $840/mo | Add as exact negative in Campaign B |
| 2 | [crm for startups] | Category campaign | PMax | $320/mo | Add as exact keyword in Search |

### Total estimated cannibalization cost: $[X]/month

### Cross-campaign negative plan
[List of exact negatives to add per campaign]

Ongoing Prevention

Cannibalization is a recurring problem in growing accounts. Prevent it from re-emerging:

When adding new campaigns

  • Before launch: cross-reference new keywords against all existing campaign keywords
  • Add negatives immediately — don't let the new campaign overlap for even one day

When expanding match types

  • Adding broad match to an existing campaign: add all exact/phrase keywords from other campaigns as negatives in the new broad match campaign

When adding PMax

  • Immediately audit for overlap with existing Search campaigns
  • Add all top converting Search keywords as exact match in Search campaigns to preserve priority

Monthly maintenance check

  • Pull search terms report, segment by campaign
  • Flag any query appearing in 2+ campaigns
  • Confirm no new exact duplicates have been added

Common Mistakes

Thinking Smart Bidding solves this Smart Bidding partially mitigates cannibalization by adjusting bids, but it doesn't prevent two campaigns from entering the same auction. You still pay the inflated CPC floor caused by your own second campaign.

Only checking exact duplicates Most cannibalization in modern accounts is match type overlap, not exact duplicates. An account can have zero exact duplicate keywords and still have severe cannibalization from broad match expansion.

Ignoring PMax PMax is aggressive about entering Search auctions. Without deliberate exact match keyword coverage in Search campaigns, PMax frequently captures queries you intended to serve with controlled Search campaigns.

Adding negatives at ad group level only If you add a keyword as a negative in one ad group but not the whole campaign, other ad groups in that campaign can still cannibalize.


Related Skills

  • google-ads-negative-keywords: Full negative keyword strategy — cross-campaign negatives are the primary tool for fixing cannibalization
  • google-ads-search-term-mining: Identify which campaigns are actually winning each query via search term analysis
  • google-ads-bidding: Smart Bidding behavior in cannibalized accounts — how the algorithm responds to competing campaigns
  • google-ads-segmentation: Campaign-level performance splits help quantify the CPA gap between competing campaigns

平台分布

Codex

36.74%
按下载量换算127

Claude

29.86%
按下载量换算103

Cursor

17.16%
按下载量换算59

Gemini CLI

10.05%
按下载量换算35

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills