Datalab Logo

March 22, 2019

Velosearch – Scraping


The internet is full of information. This information can be important or can support business decisions. The intention of this project was to extract information about bicycles from 28 different German bike selling websites. The information included the brand, model, frameset, brake types and prices listed from the different websites.

With a matching algorithm, the bikes from the different websites can be matched to each other. In this way the bikes and prices of the bikes can be compared. In this way, the prices of the different Pon brands can be monitored. Additionally, this information can be used when introducing a new bike model. A comparison can be made with other brands, to perform competitive pricing.

  • Kevin Haver

    Lead Data Scientist
  • Ellen Mik

    Data Scientist