Welcome to Meta Ads Creative Analytics
Complete these steps to get started:
Ad Accounts
| Account ID | Name | Active | Last Synced | Creatives | Actions |
|---|
Creatives
Win Rate Analysis
Loading win rate data...
Kill / Scale Recommendations
Loading recommendations...
Iteration Priorities
Loading iteration priorities...
Coverage matrix ·
Cell = batches tested (hits marked ✓). Empty cells are untested combinations — the gaps in your split.
Batch Ledger
Account Patterns
How the account is moving across time windows, winning combinations, and what to do next.
Creative Team Performance
Batches produced per week · by strategist
Weekly timeline per strategist
green = hits · grey = rest of batches · same scale on every chart · up to 26 weeksCreative Strategists
Weekly timeline per editor
green = hits · grey = rest of batches · same scale on every chart · up to 26 weeksVideo Editors
Action Queue
Task rules are managed in .
Generate Report
Generate a performance snapshot from your currently synced data window. Reports include spend, ROAS, CPA, top performing creative types, and AI-generated insights. Trend changes are compared against the previous saved report (snapshot vs snapshot), not strict calendar weeks.
Report History
| Date | Account | Spend | Avg ROAS | Avg CPA | Best Asset Type | Best Hook Tactic | Actions |
|---|
Meta (Facebook) API
How to get your Meta Access Token:
Gemini API
Get your API key from Google AI Studio
Master Creative Sheet
Set sharing to "Anyone with the link → Viewer". Auto-detected columns: BATCH #, Status, Author, Ad Concept, Avatar, Mass Desire, Awareness Level, Testing Notes, Ad Type / Creative Type (LFS, THAI, UGC…), Ad Format.
Winning batches everywhere (Testing Mix, Team, Patterns, To-Do) follow the "What counts as a winning ad?" card above — one definition, every tab.
Task Rules
Every launched batch is checked against these rules — matching batches appear as tasks in the To-Do tab. Leave a condition empty to skip it.
Strategist KPIs
Assign each strategist a weekly batch target and their lane (angles, creative types, format). The Team tab shows every strategist's live progress against these — this week's output vs target, win rate, and their assigned lane on their scorecard.
Notion — Editor Sync
Reads your Notion Tasks Tracker (tasks named "BATCH#123 — …") and fills for each batch: the Video Editor (Assignees matching the editor list below), the Strategist (from the task title / other assignees), and the card created + due dates. Exact launch date always comes from Meta (first ad's creation time). Create an internal integration at notion.so/my-integrations, share the Tasks Tracker database with it, and paste the secret here.
Team Access
Set passcodes to protect the app. Admin sees everything; team members see Creatives, Analytics, Testing Mix, Team, To-Do and Reports — but not Settings or Accounts. Leave both empty to keep the app open. (Replit Secrets ADMIN_PASSCODE / TEAM_PASSCODE override these.)
Employee accounts
Personal logins per employee. Members see Creatives, Analytics, Testing Mix, Patterns, Team, To-Do and Reports — Settings and Accounts are hidden and blocked. Admins see everything. Create your own admin account first.
Sync Mode
Data Sync
How often sync runs is controlled below (manual/daily/weekly). How much history sync pulls is controlled by the header Data Window picker.
First sync for a newly added account automatically runs a baseline import (Last 180 Days), then future syncs use your selected data window.
Automatic pull schedule from Meta. Manual sync is always available from the header.
What counts as a winning ad?
Used by Win Rate Analysis, Patterns and Kill/Scale. Unaware / Problem Aware ads are judged mainly by engagement (hook rate, CTR), Solution Aware by CTR + ROAS, and Product / Most Aware conversion ads by this ROAS threshold.
Winner = ROAS ≥ threshold AND spend ≥ min spend AND (if set) CAC ≤ max. Example: 1.0 ROAS, €300 min spend, €80 max CAC.
Minimum spend for AI recommendations
Controls which ads receive AI-generated iteration recommendations (strengths, weaknesses, and what to test next). Ads below this spend amount will be skipped since they may not have enough data for meaningful recommendations.
Set to 0 to generate recommendations for all ads regardless of spend