A retailer in the consumer electronics industry was facing a challenge in the markdown pricing process of end-of-life products. Decisions were made empirically and, although criteria had been set as guidance for the discounts to apply based on average sales and stock, most prices were set by hand as they didn’t account for many other factors, resulting in a time-consuming task that was prone to error or bias.
They felt, as a result, the need for a more consistent and reliable process of decision-making, with added time efficiency and improved data visibility. The project’s main goal was to develop an analytical model to optimize clearance pricing of discontinued items and address the challenges mentioned above.
A customized Decision Support System was developed to improve the markdown pricing process of end-of-life products. The solution relied on the implementation and integration of several components, including a machine-learning model to predict the variation in demand caused by different price reductions and a clearance pricing optimization algorithm to recommend the optimal markdown to be applied to each product.
The markdown recommendation can be automatically adjusted according with the retailer’s strategy and priorities, as it considers the margin and sellout targets set by the product managers.
Finally, the Decision Support System comprises an easy-to-use interface in which product managers validate the recommended markdowns and have access to all revelant information needed to make the best possible decisions.
The new tool allowed for accurate sellout predictions until the end of each campaign and faster reaction on end-of-life items. On average, more than 90% of the recommended price markdowns were accepted by the managers. As a result, the sellout rate increased compared to previous campaigns, without compromising margin targets
All product managers agreed that the new tool facilitated and improved lifecycle management and made the markdown pricing process much more efficient.