Skip to content

AI Predictive Model for DV360 Revenue Optimization

Development and deployment of an AI predictive model to optimize DV360 programmatic campaigns for a leading luxury e-commerce platform, predicting high-value customers from early-stage signals.

What NIMA Digital Built

  • Developed AI predictive model incorporating product views, add-to-cart, and purchase data to predict purchase likelihood
  • Implemented real-time data tracking feeding predictive signals into DV360 campaigns
  • Designed custom bidding strategies optimizing for predicted high-value customers
  • Ran A/B test with equal investment to validate the AI model's effectiveness
  • Contributed to +27% year-over-year revenue growth target for 2023

Our Approach

In collaboration with Google and Making Science, NIMA Digital's founder addressed the challenge of low conversion rates relative to high product engagement. The team built a predictive AI model that used early-stage micro-conversion signals to identify high-value customers before they purchased, enabling DV360 to optimize campaigns for predicted converters rather than relying on limited historical conversion data.

Published Results

+106% revenue increase in A/B test

Source: Making Science / Google

+100% ROAS improvement

Source: Making Science / Google

+36% sales increase

Source: Making Science / Google

-25% Cost Per Acquisition (CPA)

Source: Making Science / Google

Technology Stack

Google DV360

Programmatic display and video advertising platform

Making Science

AI predictive modeling and custom bidding

Recognition

Published Making Science & Google Case Study

Quote from Nicola Antonelli, Chief Revenue Officer

Ready to transform your digital presence?

Discuss Your Project with Our Founders

Related