Forecasting
Predictive Analytics
By analysing pricing and rankings across comparison websites, Energiedirect can predict how its competitive position relates to online sales.
Energiedirect is a Dutch energy supplier offering electricity and gas at competitive rates. As part of the E.ON Group, the company focuses on accessible services, straightforward communication and competitive energy prices.
Energiedirect wanted to understand how its position on price comparison websites affected its online sales. The goal was to identify which rankings and price differences had the greatest influence on sales and use these insights to predict future online performance.
1
Energiedirect can better understand how its position on price comparison websites influences online sales.
2
A two-phase Random Forest model turns daily pricing and ranking data into sales forecasts.
3
An interactive interface allows the team to test pricing scenarios and anticipate their impact on online performance.
A two-phase Random Forest model was developed using several data sources:
The first phase focused on analysing how rankings and price differences influenced online sales.
The second phase predicted the number of sales generated through the comparison websites on a given day. These predictions were then used to forecast Energiedirect's total online sales.
An interactive interface was built around the model, allowing Energiedirect to adjust its prices and position on comparison websites. The team can explore different scenarios and see how these changes affect predicted online sales.
.jpg)
.jpg)
.png)
