Assortment optimization:­ transform sales periods into a marketing instrument

foto van Sander Harryvan
Sander Harryvan
Marketing Manager

March 29th, 2018

Assortment optimization with Data Science has the potential to drastically change the role of the yearly sales period in the fashion industry. Traditionally, this is the time for retailers to make room for new collections and for the consumer to go bargain hunting. However, despite the lower prices the most popular items are often already sold out, or only available in the sizes XL – XXXL. The above mentioned doesn’t improve the customer satisfaction during these sales periods.

So, wouldn’t it be better for both, retailers as well as consumers, if supply and demand were more in sync? Retailers wouldn’t miss out on profits as much as they do now when selling their less popular products at lower prices and shoppers wouldn’t be so disappointed when they can only choose between oddly sized items with discount. Moreover, nowadays in shopping areas it feels like the sales period is an all-year round event with heavy price promotions everywhere and anytime. This doesn’t contribute to a positive brand image, especially for mid-market to high-end stores.

example of a traditional sales period
Example of a traditional sales period shopping environment

Let’s take a closer look at how Data Science in combination with some creative thinking could change the sales periods, resulting in a competitive advantage and more importantly happy customers!

Assortment optimization: predict demand on SKU / shop level

With the use of Data Science, it is possible to carefully manage stock levels for each SKU in each individual store by predicting the demand of every product until the end of season. Usually, retailers have a few standardized formulas that they apply to all their stores. However, don’t consumers behave differently in every shop and aren’t needs and trends location dependent? We see these differences between cities in our retail client’s shops every day. Therefore, we need to plan assortments on an individual shop level. This way we find the optimal stock level for each product at all times. This means no empty shelves (and thus no longer selling a ‘no’ to your customers) and minimizing the risk of redundant stock at the end of the season (and thus no unplanned promotions). These predictions are based on numerous factors for each specific SKU, such as product features, price elasticity, seasonal and weekday sales effects.

assortment optimization with Data Science
Assortment optimization with Data Science

Predicting the optimal assortment results in higher overall profits for retailers and enhances the customer experience for shoppers: no longer will they find their favorite products to be out-of-stock.

Check the Swarovski client case and discover how we optimized their assortment. 

Does this mean that promotions during a sales period become unnecessary? Well, no. Consumers got used to the fact they can buy retail products for promotional prices at the end of the season. It seems that it has become part of our culture. By not providing consumers the opportunity to buy at discount prices, retailers will miss out on a lot of customers in their stores. Shoppers will simply go to a competitor’s store that does have items on sale. So, an innovative approach is required to organize these sales periods, as they become theoretically obsolete because of smart assortments.

Take back the power with assortment optimization!

Applying assortment optimization enables us to plan ahead. Already before the sales period starts, we can take into account that we are going to sell a particular number of each SKU for discount prices. By carefully planning which items the retailer is willing to sell during price promotions, the retailer will take back control. Because of accurate assortment predictions ‘everything must go’ sales are ancient history. We just have to be a little bit creative on how to approach the inevitable sales season.

Thus, Data Science enables us to turn these sales periods into planned marketing campaigns. Since we can accurately predict which products are going to be sold in a particular time frame for a specific price in a specific area, we can also predict which products will not be sold until sales season and for what prices these products eventually will be sold. This give rise to the opportunity to carefully plan our assortment in these promotions periods. Hereby we make sure sufficient desired items in the right size are available in your store.

example of optimal assortment
Optimal assortment durings sales periods

By adopting a data driven assortment planning solution, customers will find their favorite items in the right size when visiting your store. It is likely, that they even went to a competing store and only found the product they were looking for in size XXXL, while they were looking for an M, resulting in disappointment. Finding the desired product greatly improves the customer experience in your store. Consequently, loyalty and customer satisfaction will increase among consumers.
This has a direct impact on profit and Customer Lifetime Value. Retailers no longer have to accept the classic rules of end-of-season sales but are able to plan them strategically to create long term value for their businesses. This will give retailers the opportunity to have a real competitive edge in the coming years.

We’ve seen that Data Science can give new meaning to classic business processes and marketing instruments. So be creative and take advantage! New technologies continuously lead to new insights. Take back control, optimize and create the opportunity to use classic instruments in an entirely different way.

The concept of improving business processes through Data Science is, of course, not limited to sales promotions. Also read about the top 4 data science trends that will have a serious impact on your business in 2018.

Got inspired and have an innovative idea, give us a call and let’s make it real together!