Demand forecasting: your purchasing needs at a glance

In today’s fast-paced world, companies and their teams must adapt to market needs to remain competitive in a complex global context. This has been further intensified by the health crisis, which will continue to test the productive fabric during the economic recovery expected for the end of this year and 2022.

In this context, as we have previously mentioned, process automation is essential to gain efficiency and profitability. New technologies provide us with support systems for business management across all business areas and departments.

The evolution and democratization of big data, artificial intelligence, and algorithms allow for the implementation of systems such as predictive maintenance, computer vision, or demand forecasting in any business environment, regardless of its size. In this post, we will focus on the latter to explore its potential, such as streamlining tasks, simplifying decision-making, increasing efficiency and profitability, and reducing uncertainty.

Future predictions

First, to provide a general overview, let’s start with the fact that demand forecasting, inventory management, and purchasing planning systems based on artificial intelligence algorithms predict how sales will behave in the future. In other words, they show us, as if we had a crystal ball, which of our references will be in demand and in what quantities over a certain period.

This information allows us to work within the real context we are going to face, enabling proper warehouse management and avoiding overstocking or stockouts—both situations that can lead to financial losses and customer churn. The former, an excess of references, leaves capital tied up, increases the risk of loss due to obsolescence, and consumes warehouse space. The latter can lead to a loss of trust from buyers who cannot be served the product they demand, at least within the desired timeframe.

In the case of our proprietary tool, Purchases—which we know well and can provide concrete examples of—it not only issues forecasts but also monitors stock levels, supplier lead times, and orders in progress. It suggests purchases automatically and in a planned manner based on these parameters and demand forecasting, allowing for the setup of customizable alerts to monitor key events and guarantee service levels.

Simple and intuitive visualization

The utility of all this information is indisputable; however, it requires the support of a friendly and simple visual environment that makes working with these parameters intuitive and fast. To maximize the value of demand forecasting systems, it is necessary to know, at a glance, what the needs of the Purchasing departments are.

In our case, visualization with Purchases is flexible and adapts to the client’s needs, thanks to an information organization standard that allows us to remain relatively independent of the chosen visualization software. We have successfully implemented the tool by coupling it with visualizations in Power BI, Qlik Sense, Tableau… although, due to price and versatility, we currently recommend Power BI.

As shown in Figure 1, we can check all the information necessary for managing a specific product at a single glance, quickly and easily: the evolution of predicted consumption in different markets; the evolution of stock status; the Predicted Average Daily Consumption (PADC), the Real Transit Time (RTT) of orders, working stock, safety stock, the days remaining until the optimal order date, and the order date itself. Furthermore, in the specific example we have chosen, a stockout warning appears, urging the Purchasing teams to acquire the queried product.

As we can see, by having a demand forecasting system and agile dashboards, work is simplified because all information is collected in one place and processed to offer a complete view, not only of the current warehouse status but also of future needs—and that is the key. Planning, therefore, moves from being based on the experience and intuition of teams to being based on certainties, which will allow operations to be adapted to real needs, minimizing time, costs, and potential losses, while maximizing profits and security.

And this is not an empty statement:

• the estimated stock reduction in the first phase of implementation exceeds 15%

• a greater negotiating power with suppliers is obtained, thanks to the forecasting of needs

• it enables the implementation of direct supply strategies from supplier to customer for those references with a higher risk of obsolescence.

• as previously mentioned, obsolescence levels and risk are reduced, since these tools allow for the detection of excess stock in low-rotation references.

• the automation of purchasing management leads to a considerable reduction in time spent analyzing planning and defining purchasing needs, which are also much more efficient.

• demand forecasting also allows for the acquisition of related items from the same supplier, optimizing the transport network and improving supply costs.

• it provides the capacity to react to variations in demand

• working with predictive analytics provides reliable (multivariable) information to improve future decision-making, both in negotiations with suppliers and for launching marketing campaigns, offers to reduce low-rotation stocks, etc.