How to Forecast New Products: Challenges and Solutions

This article tackles the challenge of forecasting new products. First, we divide product introductions into three cases: new version, line extension, and brand-new products. Then propose methods to forecast demand in each case.

Businesses like to renew their product portfolio by releasing new versions of existing products or launching brand-new products. By providing new value propositions for consumers, companies hope to increase (or secure) market share and enjoy overall sales and profit growth. Unfortunately, forecasting new products is especially challenging for demand planners, marketing teams, and statistical forecasting models. This problem is specifically serious for the fashion and IT industries as well as retailers, as they renew their products regularly.

Segmentation

Let’s tackle this challenge by dividing product introductions into three sub-cases:

• #1. New Version: New product launch replacing existing products (clear predecessors).
• #2. Line Extension: New product added on top of the current lineup (with a few similar products).
• #3. Brand-new Product: New product without predecessors.

Table 1 Examples of product launches.

Case 1: New Version (Replacing Predecessors)

Launching a new version of an existing product is the simplest case of product introduction: you can simply copy relevant data from the older product to the new one (relevant data typically includes forecasts, historical demand, and historical demand drivers such as pricing, promotions, and shortages). We also say that you use the older product’s historical data as a proxy for the new one. This will give you a good starting point for estimating the future demand for this new version.