- Watching Four Videos
- Downloading the Marketing Campaign datasets
- Before working on the business case asks yourself (and write down the answers) to the following questions:
- What Business Problem you are trying to solve?
- Why solving this problem is important?
- How did you solve this problem?
- Why would I consider your solution?
The goal of this study is to create a customer response model based on past responses to targeted marketing campaigns, in order to predict those customers that are likely to respond to and increase the exchange rate of new campaigns.
1. Step 1: Load and prepare data from past marketing campaigns, including recipient attributes (e.g. age, gender, and behavioral attributes such as the usage of products and services, website, etc).
2. Step 2: Determine which factors influence the response to market campaigns to improve prediction.
3. Step 3: Train and score the customer response model. You need to find the best model.
4. Step 4: Load data containing potential recipients for new campaigns. Apply the customer response model to identify and target those recipients that are the most likely to respond to the marketing the campaign in the desired way.
- ROC Comparison between all models you used
- Keep tuning these models to get a better Precision? Recall? …
- LIFT CHART: Create a lift chart to evaluate the ability to identify groups with a higher probability to respond to this marketing campaign.
- Consider using a Cross-validation operator.