In this project, the obtained data "marketing_data.csv" is analysed to understand the status of a marketing campaign and propose data-driven solutions to improve the marketing campaign results. The marketing data was sourced from Marketing Analytics by Jennifer Crockett on Kaggle.

Machine learning models were built to predict store purchases which have more of a continuous nature. The machine learning algorithms applied are:

Decision Tree Regressor

Random Forest Regressor

📍 The analysis showed that the Random Forest algorithm provided a better model in terms of performance compared to the Decision Tree algorithm.

📍 The insights from the analysis showed that the most recent marketing campaign "Response" generated the highest results so far. Thus more budget can be allocated to the same campaign set-up while experimenting or testing different campaign setups for more results.

📍 Based on the audience insights, more adults with the defined demography should be targeted.

📍 The highest purchases and spending capacity came from Spain. So, more campaigns can be tailored to audiences in Spain.

📍 Most customers spent more on wine and meat products. These products can be promoted more with special offers and deals to boost overall purchases. The other brand products and components of the marketing campaigns can be improved upon. However, it is important to focus more on what is working to reach more customers and boost revenue for the brand.

View the full project on GitHub - https://github.com/AdemolaI/Artificial-Intelligence-and-Machine-Learning-with-Python

References Source of Data ("marketing_data.csv")- https://www.kaggle.com/code/jennifercrockett/marketing-analytics-eda-task-final/input

Data analysis source: Marketing Analytics by Jennifer Crockett on Kaggle- https://www.kaggle.com/code/jennifercrockett/marketing-analytics-eda-task-final/notebook