Applying your data science toolkit
July 2025
Last update: 20 July 2025

The goal of this project was to predict the success of recruitment campaigns based on their setup and budget allocation.

A dataset of 446 past campaigns was used, including campaign duration, budget per advertising channel, and whether the campaign resulted in at least one qualified candidate.

The task was to clean and prepare the data, engineer meaningful features, apply classification models, and build a tool to estimate the likelihood of campaign success before launch.

Deliverables:

ML to predict the success of recruitment campaigns based on their setup and budget allocation.

The main objective of this project is to develop a classification model that can predict whether a recruitment campaign is likely to generate at least one qualified candidate. This allows marketing teams to simulate different campaign setups and budget allocations in advance, gaining insight into the expected success of a campaign before it launches.

The deliverables below include:
A jupyter notebook with the full ML pipeline, the used dataset and the written report for steakholders. Additionally I added the online prediction tool that was trained on the final ML model.

ONLINE RECRUITMENT CAMPAIGN PREDICION TOOL PROTOTYPE

With GitHub and Streamlit, I deployed the tool as a functional prototype that is now accessible to end users. This setup allows users to interact with the prediction model through a simple web interface, making it easy to test different campaign scenarios without needing any technical knowledge.

Online prediction app

Get in touch

contact

Eindhoven, Netherlands.

info@esthervanhelmont.nl