Forecast future sales from historical time-series data

This analysis requires DataCello Pro

Running a forecast requires DataCello Pro. Prepare a date field and a numeric target such as sales, with one value for each daily or monthly time point.

Parse text dates first. When transaction-level data has several rows on the same date, aggregate it to the intended time grain before analysis.

Configure the forecast

1. Select the time-series dataset, click the Analysis tab, and choose the Time Series category.
2. Select Forecast from the algorithm menu, set sales as Target, and select the Date Column.
3. Enter the number of future steps in Forecast Period.
4. Review the run checklist and click Run.

DataCello uses Prophet for this analysis. The result shows observed values, forecast values, and upper and lower bounds.

Read a range, not only a point

Use the bounds as part of the result. A wider interval farther into the future indicates greater uncertainty. Treat the center forecast as an estimate, not a committed value, when planning inventory or budgets.

Current scope

• The workflow uses a date field and one numeric target.
• You can set a horizon and inspect forecast intervals.
• Manual seasonal configuration, external regressors, multi-series batch forecasts, and scheduled refreshes are not part of the current app.