Go back to list

Strategic inventory: the challenge of single item in the fashion world

Do you find yourself stocking a store with single units of a product, understood as a combination of model-colour-size? Retailers in the fashion industry often operate according to this scheme, for various reasons. However, it always leads to serious business consequences.

Consider the following scenario: a customer visits a store where he finds a product in the model and colour that meets his expectations. He want to buy it, but it turns out that the shelf lacks the item in his size. What happens next? 

In a more favourable scenario, the customer checks the product’s availability online and orders it there. In a negative scenario, he will go to another store in search of a similar or alternative product. This is not uncommon, as in the current highly competitive market conditions, customer loyalty to fashion brands is decreasing. 

The consequence of the second scenario is lost sales, which translates into lower revenue. A similar result occurs when a customer wants to buy more than one unit of a product, but the expected quantity is not available in the store.  Such events are detrimental to the business, but it seems that the consequences of stocking stores with single units of products can be even more severe. 

Out-of-stock in store  

In the scenario where the mentioned product – a single unit in a specific model, colour, and size – is sold, the store experiences an out-of-stock situation. The computer system suggests replenishing the shelf with that one unit. The trouble is that restocking always takes time – sometimes a day, sometimes two or three days.   

When we think about the collection over the entire year and all the stores within the network, it turns out that for about 20-30% of the time when the product should be on display and thus available to customers, it is not. For 20-30% of the time, products are unavailable because they have not yet arrived from the warehouse and have not been unpacked. This means that lost sales can reach very high levels. At the same time, it is difficult to measure precisely how high they are. 

Self-fulfilling prophecy

Retailers often stock their stores for the new season based on historical sales in the previous season. This works like a self-fulfilling prophecy – the store is stocked with one unit, so it sells that one unit, even though potential sales could be higher. 

One fashion retailer conducted a test. For a month, one of the stores stocked 2-3 units of each product. When I asked about the result of this test after a month, the outcome exceeded my wildest expectations. Sales increased by 42% in that store. Without any additional marketing campaigns, without price reductions. They simply provided goods to their customers. 

Michał Koziara, Chief Operating Officer, Occubee

Sales vs. sales potential

Why do retailers stock stores with single units of products? Usually, they do not have information about the sales potential in a particular store for a specific product in a chosen colour and size. Without this knowledge, it is difficult to act differently than by stocking the store with a single unit, following the pattern of previous seasons. So, how to obtain this knowledge? Advanced analytics comes to the rescue, utilizing the latest technologies and expertise in data science, machine learning, and Big Data. 

It allows answering the question of what the sales potential of a particular product is in each store of the retail chain. It enables identifying stores where this potential is higher and where one unit of the product is sufficient. 

This happens because advanced analytical tools allow analysing not only historical sales but also current information about product availability. If a single unit of a product has been sold, and the product has not yet been replenished in the store, it is hard to expect sales during this time. 

How do we do it at Occubee? 

At Occubee, we collect and analyse sales data, as well as information about store stocks, warehouse stocks, and a range of additional variables. We look at data holistically, and thus at the entire business issue. We train our machine learning models on vast, diverse datasets, enabling them to better forecast future sales. 

Moreover, in the fashion industry, we treat models, colours, and sizes separately to identify the individual sales potential for a specific product in a given size, colour, and store. Such information, automatically generated every day, for thousands of products and hundreds of stores, allows us to answer questions like how many units of white T-shirts and in what size should be found in a specific store with high probability. 

Would you like to know more?

Subscribe to our newsletter

Newsletters are sent in email format, no more often than once a month or immediately in case of significant news/changes/educational content. More in the terms and conditions.