Key Details

• CHALLENGE: Production costs in the packaging industry change at a fast pace.
• SOLUTION: An AI-based pricing-tool suggests offer prices based on historical pricing data.
• RESULTS: A semi-automated and objective pricing tool that facilitates to reach margin and hit rate objectives.

‍

The Customer

Our customer is a major player in the European paper and packaging industry.

The production and input factor costs in the packaging industry change at a fast pace. Which is mainly due to the paper price volatility.

Balancing pricing adjustments to meet market demands without sacrificing potential margins or jeopardizing hit rate goals has been a major challenge for a long period of time.

As a result, determining the price for an end customer’s offer requires a manual case-by-case approach. This process relied solely on the sales team's experience. The pricing process was slow, error-prone and highly subjective.

In addition, managing margin goals, customer prioritization and cost-effectiveness became increasingly difficult.

The major project objective was to use Machine Learning technology to suggest prices to the sales staff, that a) maintain or increase the overall profit margin and b) yield the highest possible hit rate.

‍

The Challenge

The actual pricing process‍

Creating an offer involves several IT systems and was not handled consistently throughout the entire organization. In fact, depending on the departments involved, there were at least three to four process variations, each with its own set of sub-variations. Unfortunately, the processes were not documented and hard to uncover.

‍

Data availability

Accordingly, the historical offer data was spread over various systems and databases. The starting point for the analysis was a standard report used by the sales managers for performance tracking. Direct access to the databases was not granted, and the internal resources for data provision were scarcely available.

‍

Data quality & noise

The major project objective was to use Machine Learning technology to suggest prices to the sales staff, that a) maintain or increase the overall profit margin and b) yield the highest possible hit rate.

‍

Hosting environment

The prototype and later on the fully developed software system needed to be hosted within the customers' IT landscape to be able to have live access to the data needed. 
Due to internal procurement policy, it was not possible to use an external cloud solution. So we had to come up with an alternative solution.

‍

Tool acceptance

Due to IT-restrictions, for the moment it was not possible to fully integrate the production system into the process landscape. Therefore, we faced a challenge in convincing the sales staff to adopt the new system and overcome any hesitations or reluctance to modify their current sales process.

‍

Lack of usage transparency

Due to the workers' council restrictions, it was not allowed to implement any kind of tracking that would allow measuring employees' performance. Therefore, it was hard to measure how the tool was received by the sales staff.

‍

‍

‍

‍

‍

‍