Prediction of Digital Marketing Campaign Success Using Deep Neural Network Models

Abstrak

The rapid growth of digital advertising platforms has generated large volumes of complex and nonlinear campaign performance data, making accurate prediction of campaign success increasingly challenging. Traditional machine learning approaches often struggle to fully capture these nonlinear relationships. Therefore, this study proposes a Deep Learning approach using a Deep Neural Network (DNN) to predict the success of digital marketing campaigns based on key performance indicators such as impressions, clicks, CTR, CPC, CPM, engagement rate, and conversions.This research follows the CRISP-DM framework, including data understanding, preprocessing, model development, training, and evaluation. The dataset was obtained from digital advertising platform performance reports and processed through data cleaning, feature scaling, and train–test splitting. The proposed DNN model consists of multiple fully connected layers with ReLU activation functions and is optimized using the Adam optimizer. Model performance was evaluated using accuracy, precision, recall, F1-score, and ROC-AUC.The experimental results show that the proposed Deep Learning model achieves an accuracy of 87.6%, precision of 86.9%, recall of 85.8%, F1-score of 86.3%, and ROC-AUC of 0.91, indicating strong predictive performance. These findings demonstrate that Deep Learning effectively captures complex patterns in digital marketing data and provides reliable insights to support data-driven marketing decision-making.