Case study: Dynamic pricing

A fashion retailer has commissioned COSMO CONSULT to develop a dynamic pricing optimization model to improve the calculation of seasonal price cuts and discount dates. The projected sales figures as well as the price elasticity of the products played a special role. Today, the company uses the algorithm to calculate weekly optimal prices. Thanks to the high performance, fine price levels along with any number of discount dates are taken into account. This has also reduced the employees’ workloa

• Increasing profitability through optimal prices
• Taking into account customer-specific constraints such as final inventory or price levels in the optimization
• Automated pricing
• Analysis application to identify and validate the Key Performance Indicators (KPIs)

Dynamic pricing is one of the big levers in stationery retail industry and e-commerce. At a fashion retailer with several large stores and a wide range of products, the prices of seasonal products are reduced at predefined dates to specific price levels towards the end of the season. Setting the discount dates and discount levels used to be determined by a mix of experience and expertise. The company then commissioned COSMO CONSULT to develop a model that can flexibly and automatically set prices for all products.

As part of a workshop, the project team initially identified the special underlying conditions and operational requirements for the dynamic pricing solution. Then, it was necessary to develop a forecasting module for each individual product and product group on this basis using data mining and clustering methods. The projected sales figures as well as the price elasticity of the products played a special role in measuring the influence of price cuts on the sales of the product. The results of the data analysis formed the basis for the optimization algorithm.

In practice, the algorithm determines the optimal price of the product every week taking into account relevant parameters such as reduction costs, markdown budgets or supplier conditions. The calculated value is automatically forwarded to the merchandise management system. An expert examination by employees is possible, but not mandatory. Therefore, COSMO CONSULT provides relevant information from forecasting and optimization and visualizes it using Shiny R.

The company succeeded in significantly increasing the gross profit through dynamic pricing. Meanwhile, due to the short computing time, finer price levels and any number of discount dates could be included in the model, and the effect can then be observed or evaluated. In addition, the time required for employees to implement the pricing policy decreased. The forecast results and their visualization led to a better assessment of the further course of sales and created more transparency.