Google Ads Specialist
by
Google campaign review for ecommerce teams: join Google spend to orders and contribution margin, then write a scale, pause or reallocate decision per campaign into Notion.
Google counts a conversion the way Google counts it, across its own window, crediting campaigns by its own model, and Performance Max reports the result as a single number with the channel mix hidden inside. The weekly decision is made on that number. What it leaves out is which products each campaign actually sold and what those products earn, which for most catalogs is the difference between a campaign that should scale and one that should stop. This skill joins Google spend to your orders at line level, carries product cost and returns into contribution margin per campaign, and writes scale, pause or reallocate into Notion.
What It Does
- Pulls spend, conversions and conversion value per Google campaign.
- Joins each campaign to the products it actually sold, at order line level.
- Carries product cost and return rate into contribution margin per campaign.
- Separates brand traffic from non-brand before judging performance.
- Flags Performance Max campaigns whose result depends on brand or shopping.
- Writes scale, pause or reallocate into Notion with the reasoning attached.
Use Cases
- Judge Performance Max on margin rather than on the single number it reports.
- Stop crediting brand search with revenue that was arriving anyway.
- Find the campaign selling volume into your thinnest product margins.
- Keep a record of every scale and pause so the cycle is reviewable.
Example Session
Review this month's Google campaigns and tell me what to scale, pause or reallocate.
Seven campaigns ran. Google's ranking and the margin ranking disagree on two of them.
- Campaign | Spend | Google ROAS | Contribution | Brand share | Decision
- PMax, all products | $14,200 | 5.1 | $6,800 | 47% | Reallocate, split brand out
- Shopping, non-brand | $9,600 | 2.8 | $8,100 | 0% | Scale, plus 30%
- Search, brand | $3,400 | 11.2 | $2,900 | 100% | Hold, cap budget
- Search, competitor | $2,100 | 1.3 | Minus $400 | 0% | Pause
Performance Max looks like the best campaign in the account and nearly half its conversions are brand searches that would have converted without it. Split brand out into its own campaign and the real figure becomes visible. Non-brand shopping earns more contribution on less spend and should take the increase. The competitor campaign loses money after product cost and has done for three months.
Why This Beats Prompting It Yourself
Conversion Value vs Contribution Margin: The Hidden Risk
A generic prompt reads whatever Google reported. Ask which campaigns to scale and you get a ranking built on conversion value inside Google's window, with brand and non-brand blended together and Performance Max reported as one opaque number. The campaign at the top of that ranking is very often brand search, which is to say demand you already had.
When You Still Need The Products Behind The Clicks
A campaign is worth what the products it sold are worth. This skill states 5.1 reported ROAS against $6,800 of contribution and a 47% brand share, which is what turns a scale into a restructure. Each decision goes to Notion with its reasoning, so when the same campaign comes up next month the team can see what was decided and why.
Known Limitations
Margin per campaign needs order line detail joined to the click, so where attribution only reaches the order the skill works at that grain and labels it. Performance Max does not expose a full channel breakdown, so the brand share is inferred from query and landing data and reported as an estimate. It writes decisions to Notion and does not change budgets in the Google account.
Install It
- Download the ZIP.
It is free and there is no account to create. - Unzip it into your agent's skills folder.
Claude Code reads ~/.claude/skills/, which is hidden by default: the command in the folder block opens it. Other agents scan their own directory, so drop the same folder there instead. - Ask your agent to use it.
Restart the agent if it was already running, then it picks the skill up with no config.
skills folder
Copy
~/.claude/skills/google-ads-specialist/
SKILL.md
references/brand-split.md
reviews/
# one review per month, pushed to Notion
# macOS: create the folder and open it in Finder
mkdir -p ~/.claude/skills && open ~/.claude/skills
# Windows: paste in the Explorer address bar
%USERPROFILE%\.claude\skills\Frequently asked questions
How is this different from the Google Ads interface?
The interface ranks campaigns on conversion value inside Google's attribution. This joins the same spend to your order lines, carries product cost and returns into contribution margin, and separates brand from non-brand. Those three changes usually reorder the account.
Which agents does the skill run in?
Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision.
Why does it keep splitting brand out?
Because brand search converts at a rate no other campaign matches, for demand that mostly existed already, and blending it into an account average makes every other campaign look worse than it is. Splitting it out is usually the single most useful structural change in a Google account.
Can it really evaluate Performance Max?
Partly, and it says which part. Spend and revenue are exact. The brand share and the channel mix are inferred from query and landing data, because Google does not expose them fully. The estimate is labeled, and where it is doing heavy lifting the recommendation says so.
Does it change budgets in Google?
No. It writes the decision with its reasoning to Notion and a person applies the change. The record is half the value: most accounts scale and pause the same campaigns repeatedly without anyone noticing the pattern.
What is included with the skill?
The skill itself, the brand split reference, the monthly review template, and the Notion decision format. It is free and security scanned. We re-publish the ZIP when Google changes its reporting, so download it again if a review starts missing campaigns.
Google counts conversions the way Google counts them.
Polar joins Google spend to orders and contribution margin across 45+ sources, so scale, pause or reallocate is decided on your numbers.
<script type="application/ld+json">{"@context":"https://schema.org","@type":"FAQPage","mainEntity":[{"@type":"Question","name":"How is this different from the Google Ads interface?","acceptedAnswer":{"@type":"Answer","text":"The interface ranks campaigns on conversion value inside Google's attribution. This joins the same spend to your order lines, carries product cost and returns into contribution margin, and separates brand from non-brand. Those three changes usually reorder the account."}},{"@type":"Question","name":"Which agents does the skill run in?","acceptedAnswer":{"@type":"Answer","text":"Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision."}},{"@type":"Question","name":"Why does it keep splitting brand out?","acceptedAnswer":{"@type":"Answer","text":"Because brand search converts at a rate no other campaign matches, for demand that mostly existed already, and blending it into an account average makes every other campaign look worse than it is. Splitting it out is usually the single most useful structural change in a Google account."}},{"@type":"Question","name":"Can it really evaluate Performance Max?","acceptedAnswer":{"@type":"Answer","text":"Partly, and it says which part. Spend and revenue are exact. The brand share and the channel mix are inferred from query and landing data, because Google does not expose them fully. The estimate is labeled, and where it is doing heavy lifting the recommendation says so."}},{"@type":"Question","name":"Does it change budgets in Google?","acceptedAnswer":{"@type":"Answer","text":"No. It writes the decision with its reasoning to Notion and a person applies the change. The record is half the value: most accounts scale and pause the same campaigns repeatedly without anyone noticing the pattern."}},{"@type":"Question","name":"What is included with the skill?","acceptedAnswer":{"@type":"Answer","text":"The skill itself, the brand split reference, the monthly review template, and the Notion decision format. It is free and security scanned. We re-publish the ZIP when Google changes its reporting, so download it again if a review starts missing campaigns."}}]}</script>