Up to 60 months
The example starts with one date column and one monthly sales value, checked for gaps, duplicates, and invalid values.
Fictional deliverable preview
See the model, checks, and handoff before you buy.
This example shows how one clean monthly sales dataset becomes a transparent 12-month forecast. Every company, value, and result shown here is fictional.
Fictional data
No client result claimed
Representative structure
1. Clean input
One monthly dataset, one target metric, and explicit assumptions.
One decision metric
The model forecasts one clearly defined measure, such as monthly booked revenue, unit sales, or qualified orders.
Visible assumptions
Units, date grain, missing-data treatment, and known exclusions are documented before the model is selected.
2. Holdout validation
The selected model must beat a simple baseline on unseen months.
Seasonal comparison
A straightforward seasonal or recent-period baseline gives the selected model a meaningful benchmark.
Recent months withheld
The example hides recent observations during model fitting, then compares forecasts with the actual held-out values.
MAE, RMSE, and MAPE
Error measures remain visible in the workbook, with plain-language notes on when percentage error can mislead.
3. Decision-ready output
The dashboard separates history, validation, and the next 12 months.
Actual versus forecast
Historical actuals, the baseline, selected model, holdout boundary, and forecast horizon are clearly labeled.
Base, upside, and downside
Transparent scenario inputs let the buyer review practical planning ranges without hiding the assumptions.
Visible pass or attention state
Date continuity, numeric inputs, formula coverage, and refresh readiness receive explicit workbook checks.
4. Written handoff
The buyer receives an unlocked workbook and an update path.
Visible formulas
Inputs, calculations, selected model, dashboard, and validation remain inspectable instead of being hidden behind a black box.
Step-by-step notes
The handoff explains where to add new months, how to refresh compatible sources, and which checks to review.
Five business days
The fixed scope includes one clean dataset, one target metric, a 12-month horizon, and one revision.
This sample demonstrates format and reasoning only. Forecasts are estimates, and no particular business result is promised.