Facebook Ads Specialist
by
Meta campaign review for ecommerce teams: reconcile Meta spend to net revenue and blended CAC, then push a scale, cut or test decision per campaign into Notion with the reasoning attached.
Meta reports on Meta's terms. Its attribution window, its definition of a conversion, its view of which orders belong to which campaign, and none of it reconciles to the revenue your finance team closes on. The gap is not academic: it is the difference between a campaign being scaled and being cut. Every week a media buyer makes that call across a dozen campaigns, defends it in a meeting, and has nothing to defend it with except a screenshot of Ads Manager. This skill reconciles Meta spend to net revenue and blended CAC on one definition, decides scale, cut or test per campaign, and writes each decision into Notion with the reasoning attached.
What It Does
- Pulls Meta spend, delivery and platform-attributed results per campaign.
- Reconciles them to net revenue and blended CAC on your own definition.
- Shows the gap between platform-reported and blended performance per campaign.
- Separates a campaign that is scaling from one that is buying existing customers.
- Decides scale, cut or test per campaign with the budget change attached.
- Writes each decision into Notion with its reasoning so the next review has a record.
Use Cases
- Walk into the weekly paid review with decisions rather than an Ads Manager export.
- Show a finance lead why a 4x ROAS campaign is being cut.
- Catch the campaign whose reported performance depends entirely on returning buyers.
- Keep a written record of every scale and cut so the pattern is reviewable.
Example Session
Review this week's Meta campaigns and tell me what to scale, cut or test.
Nine campaigns ran. The platform ranking and the blended ranking disagree on three of them.
- Campaign | Spend | Meta ROAS | Blended CAC | New customer share | Decision
- Prospecting-Broad | $8,400 | 1.9 | $34 | 81% | Scale, plus 25%
- Retargeting-Site-30d | $3,100 | 4.6 | $71 | 12% | Cut to $1,200
- Advantage-Shopping | $5,600 | 2.4 | $41 | 58% | Hold
- Creative-Test-Oct | $900 | 1.4 | $52 | 74% | Keep testing
Retargeting-Site-30d is the account's best campaign on Meta's numbers and its worst on yours, because 88% of what it claims was returning customers who were coming back anyway. Cut it to a maintenance budget rather than zero, since some of that retargeting does incremental work at low spend. Prospecting-Broad is the opposite case and should take the difference.
Why This Beats Prompting It Yourself
Ads Manager vs Your Own P&L: The Hidden Risk
A generic prompt reads whatever Meta reported. Ask which campaigns to scale and you get a ranking built on Meta's attribution window, counting conversions Meta chose to claim, with returning customers included and no view of new customer share. The highest ROAS campaign in that ranking is very often retargeting, which is to say the campaign buying the customers you already had. The prompt has no way to know.
When You Still Need The Decision Written Down
A scale or a cut has to survive the meeting it is announced in. This skill states 4.6 reported ROAS against $71 of blended CAC and 12% new customer share, which is the sentence that ends the argument. Each decision is written into Notion with the reasoning, so when the same campaign comes back up in three weeks the team can see what was decided and on what basis.
Known Limitations
Blended CAC needs an order source connected alongside the Meta account, and without one the skill reports platform numbers only and labels them clearly rather than presenting them as reconciled. New customer share depends on customer identification in your store, so heavy guest checkout usage will understate it. It writes decisions to Notion and does not change budgets in Ads Manager.
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/facebook-ads-specialist/
SKILL.md
references/reconciliation.md
reviews/
# one review per week, 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 reading Ads Manager?
Ads Manager ranks campaigns on Meta's attribution. This reconciles the same spend to net revenue and blended CAC, shows the gap between the two rankings, and adds new customer share. Those three columns are usually what moves a campaign from scale to cut.
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.
What does it need connected to work?
The Meta ad account and an order source from your store. With both it reports blended CAC and new customer share. With only Meta it still produces a review, labels every figure as platform-reported, and does not pretend to have reconciled anything.
Why does it keep recommending cuts to retargeting?
Because retargeting usually shows the account's best reported ROAS while buying customers who were already returning. It is rarely cut to zero: some of it does incremental work at low spend. The recommendation is normally a reduction to maintenance, with the new customer share shown.
Does it change budgets in Meta?
No. It decides and writes the decision with its reasoning to Notion. A person applies the budget change. The record matters as much as the decision, because most accounts repeat the same scale and cut cycle on the same campaigns without noticing.
What is included with the skill?
The skill itself, the reconciliation reference, the weekly review template, and the Notion decision format. It is free and security scanned. We re-publish the ZIP when Meta changes its reporting API, so download it again if a review starts missing campaigns.
Meta's attribution window is not your P&L.
Polar reconciles Meta spend to net revenue and blended CAC on one definition, so scale, cut or test survives contact with finance.
<script type="application/ld+json">{"@context":"https://schema.org","@type":"FAQPage","mainEntity":[{"@type":"Question","name":"How is this different from reading Ads Manager?","acceptedAnswer":{"@type":"Answer","text":"Ads Manager ranks campaigns on Meta's attribution. This reconciles the same spend to net revenue and blended CAC, shows the gap between the two rankings, and adds new customer share. Those three columns are usually what moves a campaign from scale to cut."}},{"@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":"What does it need connected to work?","acceptedAnswer":{"@type":"Answer","text":"The Meta ad account and an order source from your store. With both it reports blended CAC and new customer share. With only Meta it still produces a review, labels every figure as platform-reported, and does not pretend to have reconciled anything."}},{"@type":"Question","name":"Why does it keep recommending cuts to retargeting?","acceptedAnswer":{"@type":"Answer","text":"Because retargeting usually shows the account's best reported ROAS while buying customers who were already returning. It is rarely cut to zero: some of it does incremental work at low spend. The recommendation is normally a reduction to maintenance, with the new customer share shown."}},{"@type":"Question","name":"Does it change budgets in Meta?","acceptedAnswer":{"@type":"Answer","text":"No. It decides and writes the decision with its reasoning to Notion. A person applies the budget change. The record matters as much as the decision, because most accounts repeat the same scale and cut cycle on the same campaigns without noticing."}},{"@type":"Question","name":"What is included with the skill?","acceptedAnswer":{"@type":"Answer","text":"The skill itself, the reconciliation reference, the weekly review template, and the Notion decision format. It is free and security scanned. We re-publish the ZIP when Meta changes its reporting API, so download it again if a review starts missing campaigns."}}]}</script>