> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sellergen.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Demand Forecasting

> How Atlas projects future demand, what inputs it uses, and how to read the forecast confidence range.

Atlas generates demand forecasts per SKU using historical sales velocity, seasonality patterns, and any live campaigns that are likely to increase sell-through rate.

## What goes into a forecast

Atlas combines three inputs: trailing 90-day sales velocity, year-over-year seasonality from your category, and campaign-driven demand (coordinated with Adam to account for planned budget increases or new campaign launches).

## Reading the forecast

Each SKU's forecast shows a projected units-sold range over the next 30, 60, and 90 days. The range reflects forecast uncertainty — wider ranges mean Atlas has less confidence due to volatile sales history or insufficient data.

## When forecasts update

Forecasts update automatically every 7 days or when a significant sales velocity change is detected. You can also trigger a manual forecast refresh from a SKU's detail view.