Estimating Demand for Something You Have Never Sold

When bringing in new items, how can you improve your estimate of demand for purchasing?

When you create a new product in your system, define multiple attributes for that product. An attribute could be a garment's fabric composition, or the particular wood in a furniture item, or some feature of the product, such as it air fries and bakes. There will be multiple attributes for an item. Who decides what the attributes should be? The buyer. They should know what their target customers valueing when shopping for an item.

When the decision is made to add a new item to the assortment, scan your history for all the items with similar attributes to the new one. Set the default weighting for each attribute to zero. Then ask the buyer to use their judgement to pick the most important attributes for this situation and assign non-zero weights to each attribute that matters. Using a balanced score card model, the merchandiser should calculate the resulting scores for each past item in the historical list and rank them highest to lowest. Then sit down with the buyer, review the top 3 to 5 products and decide which one is the best one to pattern after. Use its history adjusted for lost sales to define the sales curve and season total sales.

You can read a more detailed version of this article in "how to do it in detail" in our Retail and Consumer Goods Industry WIKI.


Posted by Brian Hume
17th January 2024

Back to Free Content