As the rapid growth of online sales is challenging, also your supply chain needs to stay ahead of the game. New technologies as machine learning enable you to do it. In this article we want to demonstrate that using your store inventory to fulfil online orders brings huge benefits!

Have you ever tried to make an online purchase, only to realize that your desired item was out of stock? At the same time, it is very well possible that the item was available in one of the retailer’s traditional brick-and-mortar stores close to your home. Or, did you ever wanted to have a product very soon, but your webshop offers a next day delivery? Probably you switched to another website to buy the same or a similar product. And if you are the retailer in this situation, you may not have just lost the sale, but the customer all together.

Districon is convinced that using your network inventory (including your stores) enables you to optimize the decision from what location you are going to deliver to your customer. Where do you have overstock, how fast do you need to deliver, where do you want to stimulate stock rotation, etc. As we continuously explore opportunities to improve your retail supply chain, we challenged our intern Florine van Meer to develop an algorithm that optimizes the allocation of an order to a potential fulfilment hub.

Florine successfully developed an innovative machine learning algorithm to decide from which store to fulfill online orders. She focused on cost benefits, taking into account the stock levels in each location, the expected future demand, transportation time(s) and costs, as well as costs of missed sales and obsolesce costs. Results obtained from a case-study on a fashion retailer show that significant costs savings between 6 and 14% on the total online demand fulfillment operation can be obtained. These cost savings increase even more when the percentage of online sales increases.

Next to costs benefits we foresee many uplifts from a sales and speed perspective. On other hand we realize that many retailers still face challenges to have real time and reliable store inventory data. And, off course it is a challenge to set up a fulfilment operation from store, including picking, packing and shipping, without disturbing your customers.

But, we have a taken a new step forward and proved the value. We are convinced that Omni-channel customer experience is not complete without integrated distribution networks. Florine also demonstrated how the intelligent algorithm can be broken down into straightforward business rules, in this way addressing the concern of following recommendations from a ‘black-box’.

Want to know more?

Are you also going through the transformation of integrating your distribution networks? Or would you like to know more about how we integrate intelligent algorithms and optimization into decision support systems? Get in touch with Tanya Bulavskaya for further information.