Determinants of an Effective Digital Campaign: A Study of Conversion Rates Using Facebook Data
Session Title
Social Media, Online Spaces and Literature
Faculty Mentor
Anna Romanova, Ph.D.
College
College of Business Administration
Department
Computer Science & Quantitative Methods
Abstract
For businesses running digital ads, achieving a strong conversion rate is one of the main goals in each marketing campaign. Conversion rate is a key performance metric for all Facebook ad campaigns, especially when a business aims to optimize for specific actions such as sales, subscriptions, or downloads. Regression modelling can help us better understand the underlying factors of a successful digital campaign that affect conversion rates and quantify the effects of those factors. This study uses a dataset from Kaggle that describes three different Facebook ad campaigns. It provides information about the targeted audience in terms of their interests and demographic characteristics, as well as the cost and ad performance metrics for each of the campaigns. We use individual customer characteristics, impressions, clicks and the cost metric to build a regression model for the total and approved conversion rates. We evaluate the model performance on the hold-out sample and calculate the R-squared hold-out value to ensure that our model can generalize well on unseen data.
Course Assignment
BADM 571 – Romanova
Previously Presented/Performed?
Winthrop University Showcase of Undergraduate Research and Creative Endeavors, Rock Hill, SC, April 2023.
Type of Presentation
Oral presentation
Start Date
15-4-2023 12:00 PM
Determinants of an Effective Digital Campaign: A Study of Conversion Rates Using Facebook Data
For businesses running digital ads, achieving a strong conversion rate is one of the main goals in each marketing campaign. Conversion rate is a key performance metric for all Facebook ad campaigns, especially when a business aims to optimize for specific actions such as sales, subscriptions, or downloads. Regression modelling can help us better understand the underlying factors of a successful digital campaign that affect conversion rates and quantify the effects of those factors. This study uses a dataset from Kaggle that describes three different Facebook ad campaigns. It provides information about the targeted audience in terms of their interests and demographic characteristics, as well as the cost and ad performance metrics for each of the campaigns. We use individual customer characteristics, impressions, clicks and the cost metric to build a regression model for the total and approved conversion rates. We evaluate the model performance on the hold-out sample and calculate the R-squared hold-out value to ensure that our model can generalize well on unseen data.
