Free & offline Β· Excel on Windows, Mac & web

WhichForecast

Which forecast fits your data?

A modern Microsoft Excel add-in for guided time-series forecasting. Select a series, answer a short intent question, and WhichForecast fingerprints the data, ranks candidate methods with a rolling-origin MASE backtest, and writes charts and errors back to the workbook β€” fully offline. Educational only: not investment advice.

See it in action

The task pane guides you from a selection to a ranked forecast β€” fingerprint, intent, walk-forward match, and proof charts.

A short walkthrough of the add-in β€” from selection to forecast. Drop whichforecast-demo.mp4 to replace this placeholder.

Built for planners, analysts, and students

🧭

Fingerprint first

Trend, seasonality, intermittency (ADI/CVΒ²), and gaps β€” before any method is chosen, with a transparent Candidate Policy.

πŸ†

Honest ranking

Rolling-origin MASE among candidates. MAPE is warned when zeros would lie. Forecast Value Added vs naive. Best supported among candidates β€” never β€œthe true model.”

πŸ“¦

Croston family

Spare parts and lumpy demand are first-class, unlike Excel Forecast Sheet.

πŸ”’

Offline & private

Free analysis runs entirely inside Excel. One series, no account, no raw upload.

Why stay in Excel?

Demand, inventory, weather, FX, energy β€” one engine. The wizard infers series nature. Not three separate apps.

πŸ“—

Right inside Excel

No exporting to a forecasting workstation. Write-back to WhichForecast_Results.

🧬

Method DNA + Learn

Plain-language identity cards and Student / Planner / Formal copy at every step.

πŸ§‘β€πŸŽ“

Seven languages

English, French, Spanish, German, Italian, Dutch, and Turkish in v1.

πŸ’Έ

Free & capable

Core methods plus Croston. Pro adds batch, live fetch, and advanced methods.

Same family

WhichForecast and WhichDist are sister Excel add-ins from NCU Technology. Same guided style, different job.

πŸ“ˆ

WhichForecast

Which forecast fits your data? Time-series fingerprint, method match, rolling-origin MASE β€” this site.

πŸ“Š

WhichDist

Which distribution fits your data? Guided input analysis, AICc-led ranking, and proof plots.

whichdist.com

Have feedback?

Tell us what works and what doesn’t β€” it shapes what we build next.

Share your feedback