CSV / Excel
Upload authorized aggregate data, map dates, channel spend and the KPI, and review missing or unusual values.
Check data readiness →CONSISTENT DATA COMES FIRST
Start with a data path you can use today
CSV / Excel imports are available. Direct advertising API connections are in development; automated sync requirements are confirmed before delivery.
Google Sheets
Download the sheet as CSV or Excel for import. Native authorization, scheduled pulls and automatic merging require scope confirmation.
View the file import workflow →Google / Meta / TikTok / Amazon Ads
Export aggregate platform data and prepare a file for now. Live OAuth and automated pull flows are not available from this page.
Discuss an ad connector →BigQuery / Snowflake
Aggregate spend and KPIs in your warehouse, then export a file. Direct connections, scheduling and incremental updates require assessment.
Discuss warehouse integration →REST API
Developers can start with the API reference. Confirm the origin, endpoints, authentication, limits and upload workflow against your deployment.
Read the API reference →Webhook / Segment / CDP
Event ingestion and triggered retraining require agreement on endpoints, aggregation, idempotency and job lifecycle; automatic sync is not offered as ready to use.
Discuss enterprise integration →Prepare these fields first
Use a consistent date grain, currency and KPI definition. Columns can be mapped by name; their position is not a substitute for validation.
| Field | Preparation | Why it matters |
|---|---|---|
| Date | Use YYYY-MM-DD and specify daily, weekly or monthly grain | Align spend and outcomes to the same period |
| Spend by channel | Numeric columns; distinguish missing data from genuine zero spend | Identify historical variation and channel relationships |
| Business KPI | An observed outcome such as sales or conversions, with one definition | Avoid double-counting platform-attributed sales |
| Controls | Prepare candidate price, promotion, holiday and inventory variables | Choose inclusion based on the business causal structure |
| Geography, when relevant | Use stable geo identifiers; check duplicates by geo and date | More geo rows do not create a longer time history |
Start with a file and a format check
The snippet only illustrates the format. Three rows are not enough for a production MMM analysis; history length, grain and spend variation require business-specific assessment.
date,google_search,meta,tiktok,revenue
2026-01-05,3200,2800,1500,42000
2026-01-12,3100,2600,1600,41500
2026-01-19,3400,2900,1700,43500
- Align the date range, currency, channel definitions and KPI.
- Download the preparation template and check duplicates, gaps and nonnumeric values.
- After format checks, assess modeling suitability and retain the original data version.