Statistical forecasting needs history — but a brand-new product has none, and that's exactly when a bad buy hurts most: too little and you miss the launch, too much and you're sitting on dead stock. Here's how to forecast the un-forecastable.
A new product rarely arrives out of nowhere. It usually resembles something you already sell — same category, similar price, same brand. That comparable already has a demand shape (when it peaks, when it dips). You can borrow that seasonal shape and scale it to what you expect the new item to sell. This is called analog forecasting.
Better than one analog is a cohort: blend the shapes of the several closest products so one weird item doesn't skew the launch plan. And match on category first — a new hoodie should learn from other apparel, not from a same-priced mug.
Launching your first product, or your first in a new category? Fall back, in order, to: the average seasonality of your other products in that category → your whole catalog's seasonal pattern → and finally a documented seasonal template for that category (apparel peaks in Q4, garden peaks in spring, and so on). You supply an expected first-year volume for the level; the template supplies the shape.
Real products build momentum. Assuming full run-rate from month one over-orders the launch. A launch ramp (say 40% → 70% → 100% over the first three months) models the ramp-up and keeps your opening buy realistic.
Fopsis automatically finds the best comparable products (by category, brand and price), blends their seasonality, and applies a launch ramp — falling back to category or catalog patterns when you have no comparable. You can set an expected year-one volume at product creation, or override the comparable by hand. Every new-product forecast says exactly where its shape came from.
Fopsis forecasts new products from day one — transparently.
Start Free Trial