The Opportunity
- An operator of about 1700 gas stations needed to set fuel sales volume and profit targets at the corporate level. While they used a common software solution for pricing gasoline and reporting on metrics, the tool lacked advanced analytics and optimization capabilities to support their need.
Our Approach
- We built a pricing model, leveraging their price and sales data, as well as competitive pricing, to measure price response at the store / fuel grade (e.g. unleaded) level. The response model was then fed to an optimization engine that takes into consideration margin, business rules, and other constraints. An important factor was that the company wanted to maximize overall corporate-wide volume and profit, so the model accounts for pricing at all stores collectively.
- This automated model makes daily recommendations for the pump price of each grade, and is integrated into the electronic signage outside each store. Humans no longer need to manually change prices in the system.
- Ancillary analytics include insights delivered regarding the tradeoff between volume and profit, understanding of the unique market for each store, competitive behavior, and the segmentation of stores.
The Impact
- A pilot test showed an immediate impact on profit such that the CEO ordered a full implementation be immediately undertaken at an accelerated pace. On two public earnings calls with investors, the CEO touted this initiative as one of their top strategic programs to drive corporate performance.
- Both in pilot and in real-world evaluations, the system was shown to increase profits by 10%.