Our client, a large retailer, needed help with winning back its lost customers through their year-round marketing initiatives and achieve its membership acquisition goal for the year.
The client wanted to establish reliable, scalable and impactful marketing strategies to better understand and engage with this target group of customers.
We first identified members who had left in the last two years and were suitable for Win Back Campaigns.
We divided the target group into multiple segments primarily based on their delete date and engagement while they were members.
Multiple models were built to test effectiveness using historical data as a training set and deploy an Ensemble model for scoring members.
The model was then used to develop an experimental design with all the relevant inputs, which could be used to track campaign performance across channels. This helped measure lift across channels and evaluate offers to assess what worked and what needed to be improved.
In Direct Mail, the lift calculated (Test vs Control) was close to 3%, significantly higher compared to previous initiatives.
The offer with extended membership (Trial Period) was more impactful compared to providing money off on products.
The Win Back members acquired had a High Lifetime Value compared to the previous iteration.
These recommendations were shared for model improvement and optimisation for the future.