There is a saying among forecast users: “If you forecast, you may be wrong; but you will always be wrong if you do not forecast.”

Demand planning is crucial for companies to determine the figures they need in different areas of operations, such as production planning, purchasing raw materials and inventory investments. But how much faith are companies willing to place in inaccurate forecasts?

Many companies use demand forecasts that are based solely on historical data. This might be good enough for products that behave consistently over time and in markets that remain stable. However, these forecasts are less effective in a volatile global market facing sociopolitical challenges, such as trade wars and Brexit. Additionally, increased demand for product variability compounds forecast complexity. With these situations, timely and accurate insights are crucial in making the right decisions for your business – but how do you plan for the future while facing uncertainties in the present?

In such circumstances it is better to have a range forecast instead of a single value forecast.  A range forecast – otherwise known as a prediction interval – is  an interval with a specified probability that the actual demand will lie within the interval.  A 90% interval identifies the range for a demand in a specific future time period.

Complexity in a product portfolio – which can consist of hundreds of products – makes creating an accurate forecast a challenging task. Furthermore, the effects of new items or services on the demand of existing products (e.g. cannibalization) have to be considered. Discounts and sales channel or market-specific promotion campaigns also affect forecasting. This combination of factors creates too many variables for a single team to create detailed forecasts for every product.

While forecasting aims to determine future demands, trends are not stable over time. To deal with trend fluctuations, your forecasting solution must be able to generate a number of scenarios. Millions of pieces of information must be organized, reviewed and calculated. This is impossible to do manually. Information is scattered, with different departments possessing specialized knowledge and access to different types of data. Without a sophisticated tool that supports complex forecasting processes, your forecasts are less accurate and lack effectiveness. For better forecasting and decision-making support, you need a tool that includes:

• Algorithms that automatically ‘clean’ your input data, ensuring that the numbers are accurate, consistent and complete
• A set of unique machine-learning algorithms that enable the system to recognize value in data and identify trends to continuously improve on forecasting (for example, the Management by Exception process which identifies items that meet a set of exception criteria defined by each company’s operations group)
• Real-time collaboration between departments in your company, resulting in a consensus forecast

Another key feature to consider when selecting a forecasting tool is whether it enables users to collaborate by providing visibility, enabling input from multiple sources and aggregating information to

reach a consensus and generate a single forecast – the consensus forecast.

The DELMIA Quintiq Demand Planner is capable of pulling together fragmented data from different departments and streamlining the forecasting process by eliminating time-consuming, manual tasks in the planning process. This ultimately improves the quality of the forecasts. In addition, it helps you to cluster products according to  the customer’s criteria. You can even opt to implement automated forecast planning on lower priority products. Whether it’s the launch of a new product or service, a discount or a promotion, the DELMIA Quintiq Demand Planner incorporates this information into its calculations.

Our Demand Planner solution uses a comprehensive library of state-of-the-art algorithmic statistical forecasting models. It has an Automated Model Selection feature, which identifies the most appropriate forecast model based on the historical data characteristics of the item. Fronting the solution is a user-friendly interface that does not require a background in statistics for its use.

The DELMIA Quintiq Demand Planner also provides an in-built functionality that enables planners to not only create and manage different sales events, but to see the impact of these events on other forecasts. The forecast results can also be easily derived, so planners can immediately see if the forecasts meet, surpass or fall short of their target sales, inventory, and other KPIs.

Learn more about the DELMIA Quintiq Demand Planner on our website.