With 6 months in advance, the commercial time of a large retail chain needs to buy the products of the next collection, to do this they have a poll where everyone votes on the products they believe in the most. The team has around BRL 300 million to buy around 1000 available products. With analyzed only in team insights, the stockout of Best Sellers products, as well as a large amount of stopped products (without successful sales) becomes a routine. In addition, there was little data to build a model that could define how much to allocate per product, considering that all products were new and without history.
From the described scenario, it was possible to identify inefficiency in the allocation of resources, because if the choices are wrong and there are parts of little-sold products or there is a stockout of Best Sellers products, it causes the loss of millions in investment. Thus, a model was thought of to classify the products that are the best sellers in the collection. Knowing the Best Sellers, the quality distribution of these products is carried out, avoiding problems of lack of stock or surplus, buying the items in the ideal quantity.
Our team developed a powerful tool that not only shows you the inner workings of a complex system but also pinpoints the exact spots where the costs are piling up. That’s where the Sankey chart comes in – a visualization marvel that helps you identify the main flows and transfers within your system boundaries, and does so in a visually striking way.
Using Qlik, you can create a Sankey chart that shows the width of the arrows proportionally to the amount of flow, highlighting the stations that impact your costs the most. But that’s not all – this chart is continuously updated by an AI algorithm that uses Java language and is paired with a downtime cost KPI that provides valuable insights into the system’s health. And when critical interruptions occur, you can quickly act upon the information provided by the application, minimizing productivity loss and taking corrective maintenance measures before it’s too late.
Thanks to the project’s implementation, the client’s managers reported an impressive 11% reduction in losses, resulting in a staggering more than USD 101 thousand dollars per year, per activity. But that’s not all – the project also played a significant role in helping the company move towards achieving the World Class Manufacturing (WCM) certification, specifically in the cost management pillar. The algorithm utilized in the project received glowing evaluations from the certification body, confirming the project’s effectiveness. Furthermore, the innovation contributed to standardizing data from different plants, making it easier for teams to share information and improve communication.