Datalab Logo

July 1, 2021

PAH – Network Management Tool

For the network strategy team of Pon Automotive we created a new, interactive tool which helps examine the dealer networks of all car brands of Pon. The tool allows for a full overview of the current network, scenario analysis for changes in the network, and a deep dive to zoom in to the deepest level and see where each dealer performs best.

To start measuring performance of dealer locations we first assign each postal code in the Netherlands a potential. This is done for both sales and service, and is based on the actual registrations in the area. When the registrations were not accurately put in a postal code, we used an algorithm to spread the potential to areas with similar characteristics. Combining this with a drive time curve gives us the ability to assign a sales and service potential to all dealers.

Each dealer is assigned potential based on the coverage of surrounding reachable postal codes. The potential is an indication of the sales or service a dealer is able to do based on its location with surrounding areas and competition.

The total score of the network is calculated based on all dealer locations and all assigned potential to dealers. This also shows us the total coverage in the network and in which areas we do not yet have an optimal coverage. This can be analyzed on a dealer, organization or market area level to see which have the best performance.

The most exciting feature in the tool is the possibility to quantify exactly the influence of changing a dealer location, by adding or removing a dealer from the network. On all different levels we check the influences of the changes in the scenario. This allows the network strategy team to make data driven decisions for actions in the dealer network.

The tool can also be used to show dealers where they are optimally covering an area, or where there is still potential available to increase sales or service. It also displays which areas the competition is fierce, and in which areas there are only a few dealers to choose from. This gives a good overview of the dealer performance and where changes in the network are necessary.

Finally, we can dive into a scenario and see exactly where each dealer gets its potential from, and who are the competitors for each area. This is made as transparant as possible, to give the network management team the best possible tools to make optimal decisions for the dealer network.

  • Kevin Haver

    Lead Data Scientist
  • Jesper Slik

    PhD Data Scientist
  • Wilte Falkena

    Data Scientist