Charlotte Tilbury is an internationally renowned makeup and skincare brand who wanted  to better use their existing investment in mar-tech to drive ROAS improvements across their international display channels. Simple first-party audiences like basket abandoners showed great promise but they wanted to explore “what is possible” when using ML to build these audiences from all the available signals in Google Analytics 360.

Incubeta was tasked with increasing campaign ROAS KPIs, activating effective First-Party audience segments in DV360 and demonstrating and quantifying the value that Data Science can bring when activating large volumes of first party data 

To reach our objectives we built a Machine Learning model to ingest key indicators of purchase intent from GA360 and predict the purchase propensity of each user. From here we created Audience Segments based on users’ purchase propensity pushing segments into DV360 to action more sophisticated retargeting strategies.