Case study

How a chemical manufacturer increased their market forecast accuracy by 42%

Download full case study

Their challenges

Allocate resources efficiently

With a branch in this manufacturer serving the consumer goods market, they needed to get insight into the textile production volumes and export numbers to determine the key drivers that would impact their chemical production, and guide their market strategies for both product and region categories.Not being able to forecast demand accurately and the lack of clarity over the impact of macroeconomic factors meant that they did not have full visibility over resource allocation. This led to unnecessary costs and potential lost revenue.

Effect of Macroeconomic factors

Currently utilizing a bottom-up approach, their forecasting method relied solely on a simple univariate model, which presented a significant limitation. This approach hindered their ability to incorporate macroeconomic factors or leading indicators into their aggregated forecast. As a result, they missed critical insights, as univariate models are inherently unable to account for broader economic conditions, adversely affecting the overall forecasting accuracy.

Key results

+42% forecast accuracy improvement

By implementing best practices through all stages of the forecast process, the manufacturer achieved a double-digit MAPE forecast accuracy improvement.

Identified predictive market drivers

After identifying their leading indicators and demand, they could now able to able to factor in the economic developments and business cycle impacting the specific product group. This meant that they could get visibility over the significance of each indicator and an overview of their impact on their business projections.

Overview of their market drivers at an industry and region level

The organization was now able to create forecasts easily for each product group and region. With the models built, it simply requires a quick update monthly, making the process repeatable, saving valuable time Identify their leading indicators and demand more accurately, and get visibility over the significance of each indicator.

The risks associated with only applying univariate forecasting models meant that they were missing out on the opportunity to apply leading indicators to their aggregated forecast as univariate forecasting models do not allow for that.

Download case study

Resources