/Marketing-Campaign-Analysis-with-Explanatory-Model-Analysis

This is my first project as a Business Intelligence that requires me to provide recommendations for and review data on customer feedback on previous marketing methods. One of the most needed insights in this analysis is which marketing methods can be repeated in order to make customers return to shop at our company.

Primary LanguageJupyter Notebook

Marketing-Campaign-Analysis-with-Explanatory-Model-Analysis

Have you ever gotten a shopping promo with a big discount that coincided on a certain day? Let's say that every Friday the "BaBa" shopping app gives a 50% discount for 1 month. But when you check the next month the discount doesn't exist.

Have you ever thought about what happened so that the online shopping application "BaBa" did not provide discounts again for this month? Or is your hypothesis only limited to, "maybe the company is losing money"?

This hypothesis is not entirely true. The online shopping application "BaBa" may have reviewed the campaign they did a month ago and received information that the discount shopping campaign was not able to attract as many customers as previously predicted. Therefore "BaBa" did not continue the campaign.

  • Analyzed the prior marketing campaigns of Shopping company using various ML techniques like Logistic Regression, Random Forests,Decision Trees, and XGBoost and predicted if the user will accepted the campaign term or not.
  • Recommended, the marketing team, ways to stop do the first campaign because of a lot customer did not accepted it.

There are many factors that influence the company not to continue the campaign. For example, in the following marketing analysis, it was found that the campaign that was carried out could not attract customers or even meet the estimated number of transactions.

Quick result

Feature Importance

Alt text

Campaign Result

Alt text