Static Cloudflare Pages app

Total Collected Sales Forecaster

Upload a Square transaction CSV and forecast daily Total Collected from today forward through any end date. The app ignores sales dated on or after the forecast start date, builds a complete historical daily calendar through the prior day, keeps missing transaction dates as $0 sales, and learns recurring no-sales days such as closed Sundays. Optional external factor CSVs, such as daily temperature or event traffic, are tested as residual adjustments after the calendar baseline is built, so they can improve accuracy without creating runaway long-range forecasts.

Local only Files are processed in your browser. No server, database, API call, package, or external dependency is used.

1. Upload sales data

Expected Square format: a CSV with Date and Total Collected. The app also auto-detects common date and sales column names.

Choose sales CSV No file selected

2. Optional external factors

Upload one CSV with Date + up to 3 numeric factor columns, or upload up to 3 separate CSVs. Future factor dates are used when available; missing future values are imputed from recent weekday patterns. Metadata columns such as index, status, method, and day name are ignored automatically.

Choose external factor CSVs None selected

External factor CSV format

Date,Temperature,FootTraffic,EventScore
2026-04-01,58,240,0
2026-04-02,63,255,1
2026-04-03,70,310,0

3. Forecast settings

Forecast start date Sales dated on or after this date are ignored and forecasted instead. Forecast end date Target column Date column

Include missing calendar days as $0 sales Learn recurring $0 days, such as closed Sundays Use completed transactions when status column exists

Actual Total Collected $0

Forecast Remaining $0

Projected Total $0

Backtested Error —

Forecast chart

Daily Total Collected. The shaded area is the estimated 80% interval from backtested residuals.

Model details

Daily forecast

Date	Day	Type	Actual	Forecast	Low	High	Calendar rule