Improving reach curve and budget optimisation modelling for a media agency

Media agency

We created an ensemble campaign model to improve strategic decision-making.

Background

Many media organisations struggle with incomplete or inconsistent data when it comes to assessing their campaign performance. Our client had valuable datasets, but like others in the industry, there wasn’t enough accurate data needed for making strategic business decisions. This was impacting their ability to achieve their marketing and financial goals. They needed a smart solution to process their data more effectively, turning it into clear insights that could help optimise their campaigns.

Solution

To help, we created an ensemble model, a machine learning technique that combines multiple model predictions to improve accuracy. Key elements of the solution include applying weighting to each model for specific scenarios and requirements. We considered how to address issues such as diminishing returns and data deduplication, and incorporated new data to ensure the system remains adaptive and current.

Impact

Our model provided a data driven method to make more accurate estimates of reach curves and budget allocations for our client’s customers. The media agency now has modular infrastructure in place from which they can use to optimise the solution over time, making smarter spending decision to ensure media campaigns reach as many relevant people as possible and to maximise ROI.

background placeholder
Innovation

“Our customer recognised that machine learning methods represented an opportunity to innovate within their industry. We’re delighted to be the partner they chose to innovate with.”

Samuel Crump

Engagement Manager at Faculty

partner.