Net Conversion uses cluster analysis to separate Portillo’s consumers by behavioral, psychogrpahic, and demographic traits for more targeted marketing campaigns.
With restaurants in unique markets across the U.S., Portillo’s needed an effective method of comparing relative restaurant performance as well as discerning paid media impact on total performance.
To establish subsets of restaurants to be used for more accurate benchmarking.
The main KPIs are based on driving online orders, be that delivery or pick-up. However, we wanted to gain an understanding of our impact on total performance – including offline. To measure performance, we first created clusters of similar restaurants based on seasonality, the age of the restaurant, and restaurant type/menu offerings. We then served media to a radius around a few restaurants from that cluster. The restaurants to which we didn’t serve media acted as a control, allowing us to measure the lift at the exposed store.
By identifying subsets of observations that shared similar traits including behavioral, psychographic, and demographic, we were able to bucket subsets into their own clusters for comparable benchmarks, tailored strategies, and more personalized communications.