Quality score fixer
12 free, installable Google Ads skills for Claude and other coding agents.
npx -y skills add kelpi-ai/google-ads-skills --skill quality-score-fixerAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 24 days oldThe repository was created 24 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use Google Ads Quality Score components to find the weakest link behind a valuable keyword and draft the smallest useful test. Use when relevance, CTR, or landing-page experience may be holding Search back.
SKILL.md
3.2 KB, 612 tokens by cl100k_base, as published. Nobody here has run it
Quality Score Fixer
Doctrine
Quality Score is a diagnostic, not a KPI and not a direct auction input. The useful part is the three component ratings: expected click-through rate, ad relevance, and landing-page experience. Diagnose the component, weigh it against real business performance, and test the smallest credible fix. Do not chase a prettier 1-to-10 number at the expense of conversions.
When to use
- When valuable keywords show below-average Quality Score components or weak Search performance.
- Use a connected Google Ads MCP, or paste keyword metrics, component ratings, triggered queries, current ads, and landing-page copy.
- Bring the primary conversion, target CPA or ROAS, conversion delay, and downstream quality where available.
Run it
Run a READ-ONLY Quality Score diagnosis. Use live Google Ads data if connected; otherwise use what I paste. Change nothing.
1. Confirm the date range, primary conversion, target, and keyword-level spend, clicks, conversions, value, Quality Score, expected CTR, ad relevance, and landing-page experience.
2. Prioritize commercially important keywords with below-average components. Do not rank work by the 1-to-10 score alone.
3. For each diagnosis show: weak component, business importance, observed evidence, competing evidence, root-cause hypothesis, confidence, smallest useful test, and success measure.
4. Use triggered queries, current ad assets, and page copy to test message match before proposing a rewrite.
5. Keep fixes component-specific:
- expected CTR: query fit and distinct, credible ad assets;
- ad relevance: tighter alignment among query, keyword, ad group, and ad promise;
- landing-page experience: message match, usefulness, usability, speed evidence, and one clear next step.
6. Measure both the component rating and business results after enough data and conversion delay. A component can improve while CPA gets worse.
7. If asked for savings, say they cannot be estimated from Quality Score alone. Only calculate a scenario from user-supplied CPC, traffic, conversion, and validated assumptions, labeled as a scenario rather than a forecast.
Do not change bids, keywords, ads, or pages. Do not promise a future score, rank, CPC reduction, or savings. End with "No changes were made."
Guardrails
- Quality Score helps diagnose user experience. It is not a business outcome and the displayed score is not a direct auction input.
- Never pause a profitable or strategically valuable keyword only because its score is low.
- A below-average component points to where to investigate, not the exact root cause.
- Do not claim that moving from one score to another cuts CPC by a fixed percentage.
- Fix message and experience problems before treating a higher bid as the answer.
Good output looks like
A short list of component-level diagnoses tied to valuable keywords, each with evidence, one test, and an honest success measure. No fake savings forecast and no bid change.