Forecast Preferences

Forecast Preferences are shop-wide settings that change how Logistified computes averages and forecasts. They live at Settings → Forecast Preferences and affect every forecast view, variant page, and report that runs after you save — existing saved views are not reset.

On this page

• History Window Divisor
• When changes take effect
• See also

History Window Divisor

When you ask for “the last 90 days” of sales but a product only has 70 days of data, something has to fill the remaining 20 days. The History Window Divisor controls what the average is divided by. It affects:

• Forecast models that use a sales rate (Average, Trend, and most others).
• Days of stockDays of stockHow many days your current on-hand inventory will last at your current sales rate, ignoring anything incoming. A pure runway number. Read more → and DOHDays on HandHow many days of demand your current stock covers, at the current sales rate. Lower DOH means stock is moving quickly; higher DOH means you may be over-stocked. Computed as `Average Inventory / COGS × 365`. Read more → calculations.
• Average daily and monthly sales columns.
• Reorder point and suggested order quantity math.

Match available history (default)

Divide by the number of days the product actually has data for. A product with 70 days of history is averaged over 70 days even when you asked for 90.

Use this when: you want forecasts for new or recently-listed products to reflect their actual selling rate, not an artificially low one caused by days with no data.

You ask for	Product has	Divisor used
90 days	90 days	90
90 days	70 days	70
90 days	10 days	10

Use the exact requested window (strict)

Always divide by the full requested window, treating days before the product had data as zero-sales days.

Use this when: you want every product benchmarked against the same denominator, so a new product’s rate is naturally lower than an established product’s rate over the same window.

You ask for	Product has	Divisor used
90 days	90 days	90
90 days	70 days	90
90 days	10 days	90

Note

Neither mode is universally “correct” — the right choice depends on how you think about new-product rates in your business. Default (match available history) keeps new products from looking artificially slow. Strict keeps the denominator consistent across your catalog.

When changes take effect

Changes apply the next time a forecast view, variant page, or report is computed — there is no need to reload or re-save views. This is a shop-wide setting; it does not change the history window you selected, only the divisor used in the average.

See also

• History period — choosing the window itself (how far back to look).
• Forecast models — which models use the sales-rate average.
• Reorder suggestions — how the average feeds suggested order quantities.

Last updated: 20 June 2026