Meta Ads Deep Performance Report

What it does

Pulls multiple metrics from the Meta API (placement, device, demographics, creative, destination), detects performance patterns, and produces a report with prioritized recommendations and estimated impact, using the configured conversion source as the primary truth.

When to use it

• You need a full diagnostic of why a Meta Ads account is over- or under-performing.
• A client wants a monthly report with actionable findings, not just raw numbers.
• You suspect a specific segment (device, placement, demographic) is dragging down performance.

Steps

1. Confirm the period, accounts, and primary conversion source for the report.
2. Pull metrics per campaign, ad set, and ad along with the requested breakdowns.
3. Analyze each breakdown for concentrations of inefficient or high-performing spend.
4. Detect cross-cutting patterns (e.g., a creative format performing differently by device).
5. Draft prioritized recommendations with estimated impact of implementing them.
6. Present the full report as a draft before sharing it with the client.
7. If reconciling with an external attribution source, add it as a clearly labeled optional section.

The skill

---
name: meta-ads-report
description: Deep Meta Ads performance report with placement, device, demographic, and creative breakdowns, pattern detection, and prioritized recommendations.
---

# Meta Ads Deep Performance Report

## 1. Purpose
Turn raw Meta API data into an actionable report explaining why the account performs the way it does and what to do about it.

## 2. Data scope
Define the period, accounts, and primary conversion source before pulling data.

## 3. Breakdowns
Analyze by placement, device, demographic, creative, and destination to find where efficient or inefficient spend concentrates.

## 4. Pattern detection
Cross breakdowns against each other, e.g. creative vs. device, to surface findings not visible in any single table.

## 5. Recommendations
Prioritize suggested actions by estimated impact and implementation effort.

## 6. Optional reconciliation
If an external attribution source exists, include it as a separate, clearly labeled section rather than mixing it with native Meta numbers.

## 7. Delivery
Always present the report as a draft for review before sending it to the client or final stakeholder.

## 8. GoHighLevel angle
Link creative and audience findings to GHL pipeline reporting to show the lead's full journey, not just the click.

For Claude Code, save it as ~/.claude/skills/meta-ads-report/SKILL.md. In Cursor, ChatGPT or another assistant, paste it as project instructions.