Situation

Our client, one of the world’s largest food companies, was spending a significant amount on trade promotions. It relied on a traditional parametric model, which limited its ability to deliver the right promotion at the right time and led to a loss of potential revenue.

cs-FoodManufacturerAI-hero

Our client needed to move more quickly and enable non-technical users to generate their own reports.

The company needed to move more quickly and include more data inputs so that even non-technical users could generate their own reports.


Action

We deployed the world’s first Trade Promotion Intelligence (TPI) solution, which uses AI and machine learning (ML) to improve promotion selection and timing. Having AI at the solution’s core allowed our client to include additional data sources, resulting in far more accurate models.

We used an optimization engine to analyze customized business constraints and generate calendar options to maximize sales in all feasible scenarios, and then added a machine-learning layer to continuously improve performance. Next, we developed an interactive user interface to put new insights directly into the hands of decision-makers.

This simple, intuitive solution lets non-technical users easily model complex scenarios, identify new opportunities, and preview the expected results from these actions.

Results

Now our client can generate detailed, goal-based recommendations for its retail partners in minutes.

Our AI- and ML-enhanced solution gives the company a fuller, more accurate picture and lets it make better decisions while keeping pace with the market.

Its modeling error has shrunk from about 20 percent to less than 5 percent.

The company can also make promotion decisions faster, thanks to real-time optimized calendar generation.

  • 4.7%

    uplift biggest US retailer
  • 7%

    uplift second-biggest US retailer
  • <5%

    modeling error