An Innovative Approach to Advertising Campaign Effectiveness Analysis: Traditional vs. Online

In today’s rapidly evolving marketing landscape, the ability to accurately measure the effectiveness of advertising campaigns is of critical importance. Addressing this challenge, a team of researchers led by Krzysztof Król, Tomasz Sidor, Anna Wiśniewska, Edward Józefacki, and Bartosz Bartnik has developed an innovative methodology. Their research presents a comprehensive comparative analysis of campaigns carried out through both traditional and digital communication channels, combining statistical methods with advanced machine learning techniques.

One of the team’s most significant achievements was the integration of data from diverse sources—ranging from offline campaigns (e.g., local promotions in restaurants) to those conducted on social media platforms (e.g., Facebook Ads). For traditional channels, the primary metric of effectiveness was direct sales value, whereas online campaigns were assessed using indicators such as the number of impressions, clicks, reach, and audience engagement levels.

The team applied a broad range of machine learning algorithms, including decision trees, random forest, linear regression, k-nearest neighbors (KNN), support vector machines (SVM), and LightGBM. In terms of prediction accuracy, the best results were achieved with the Random Forest model, which yielded the lowest root mean square error (RMSE = 0.385), confirming its high effectiveness in analyzing complex marketing data.

The study also identified the most influential factors affecting the performance of online campaigns. Among them were cost per click (CPC), the number of ad impressions, and contextual data such as the advertiser account name. External factors were also taken into account—including weather conditions, socio-cultural events, and the pandemic situation—which proved important in interpreting variations in campaign effectiveness over time.

The analysis was enriched with detailed data visualizations, allowing for intuitive comparison of various marketing strategies. For instance, box plots showing sales results by campaign type helped better understand data distributions and the impact of different factors on business outcomes.

The study’s conclusions clearly indicate that integrating data from traditional and digital channels significantly enhances forecasting accuracy and enables more effective advertising budget management. The proposed methodology can be successfully implemented by companies seeking precise and flexible tools for analyzing the effectiveness of marketing campaigns.

The full version of the article is available at: link to publication