Optimization of Financial Resource Allocation for Paid Social Media Campaigns Using Data Driven Approaches

Abstract

Currently, decisions on financial allocation for social media campaigns are not fully datadriven and lack precision in correlating performance with geographic trends. This results in suboptimal use of marketing budgets, reducing return-on-investment (ROI), number of leads and campaign effectiveness. The main challenges in solving this issue lie in data processing due to the complex nature of the marketing data available, selecting the correct data to enable focusing on the core problem and homogenisation of the available data sources, since several different social media platforms deliver different data. Three pillars describe the basis for this thesis. First, the role and importance of social media data will be highlighted. Next, the data will be processed, selected and homogenised. Finally, a predictive data analysis is issued. This case study has found that the data quality currently available at CAD+T is not sufficient for proper analysis in marketing analytics and reporting. Steps should be taken to allow for a homogenised data base with comparable information, like enabling a drill-down into data on a national basis rather than a larger regional basis. At least a national drill-down on all sources for digital marketing data allows for comparable data which provides deeper insights than currently possible. These insights might enable a more automated approach to budget allocation of social media budgets. Data-based budget allocation could greatly improve the ROI on spending in digital marketing efforts, allow for more concise strategies for digital marketing campaigns and potentially improve lead generation. Additionally, it could provide a marketing department, a team with traditionally hard to monitor ROI, a means of monitoring development of digital campaigns. This could allow managerial teams easier decision-making in budgeting. However, the most insightful potential for marketing and the entire company lies in the data currently not available in sufficient quality. Enabling a higher level of quality data for regular monitoring could unlock a much greater potential for insights into the ad campaigns, customers and the whole market. The specific results and the recommendation are only applicable to the company of CAD+T. Due to the custom data set from the company, there is no general applicability. The methodology, however, can be applied to companies globally which use regionally targeted advertisements. Even though this case study focussed on a B2B environment, the methodology could also be applied to B2C-focussed companies. However, in that situation, different considerations regarding qualified leads would need to be done.

Date of Award	2025
Original language	English
Supervisor	Christina Feilmayr (Supervisor)