Ad Scatter
Free Performance Analysis Tools
Analyze Meta Ads performance and identify ads that are statistically ready to scale, still gathering data, or likely underperforming.
Upload a performance CSV, choose what you want to analyze, and identify statistically meaningful Scale, Wait, and Kill zones.
Upload a Performance CSV
Drag and drop your CSV here, or choose a file from your computer.
CSV only · Maximum file size 30 MB · Processed privately in your browser.
Required: one Analyze By field, Spend, and Conversions. Optional: Campaign, Ad Set, Country, Date, ID, and Revenue / Conversion Value.
Download sample CSVMap CSV columns
The setup is organized by purpose. Choose what each graph point represents, map the required performance metrics, then add any optional context and row-level filters.
Choose What to Analyze
This grouping decision defines what each point and result represents.
Examples: Ad Name, Campaign, Ad Set, Country, Creative, or another dimension from the CSV.
Required Performance Metrics
These numeric columns power the CPA and statistical calculations.
Actual amount spent, not budget, spend limit, CPA, CPC, or CPM.
Purchases, leads, applications, or another numeric result.
Optional Context Mapping
These columns add campaign structure to results and create optional data filters.
Optional. Used for context, result columns, and campaign filtering.
Optional. Used for context, result columns, and ad set filtering.
Filter Mapping
These columns create filters that remove source rows before grouping and recalculating the analysis.
Optional. Enables a country filter on the analysis screen.
Optional. Map a row-level Date or Day column. Reporting Starts / Ends usually describe the whole export and should not be used to filter individual rows.
Advanced optional mappings
Optional. Keeps same-name items separate when the platform provides a unique ID. It does not change the Scale / Wait / Kill formula.
Optional. Used only to calculate ROAS. Do not map an existing ROAS column.
Mapped Data Preview
Performance analysis
Analysis settings
These controls tune the statistical model. They do not remove source rows.
Your acceptable cost for one conversion.
Lower confidence classifies sooner; higher confidence requires more data.
Grouped results below this spend are excluded from the chart, totals, results, and export.
Optional display cap. It does not change classifications.
Data filters
These filters reduce the CSV rows first, then regroup and recalculate every result dynamically.
Keyword Filters
Include or exclude source rows using the selected Analyze By value.
Keeps source rows whose selected Analyze By value contains any listed keyword.
Removes source rows whose selected Analyze By value contains any listed keyword.
Mapped Data Filters
These options appear only when the corresponding CSV columns are mapped.
Scale & Kill · Poisson confidence
Which entities are ready to scale?
| Point | Status | Analyze By | Spend X-axis |
CPA Y-axis |
Conversions | Expected conversions | Scale CPA Threshold | Kill CPA Threshold | Campaign | Ad Set / Group | Country |
|---|
How the model reads the data
Expected conversions are calculated as Spend ÷ Target CPA. The selected confidence level creates upper and lower conversion boundaries using a normal approximation to Poisson variation. Values above the upper boundary are labeled Scale, values below the lower boundary are labeled Kill, and the middle remains Wait.
This tool provides statistical guidance only. It does not make changes in any advertising platform.