Publicação
Taking targeting offline : targeted advertising in the real world
| Resumo: | Companies all over the world spend billions of dollars every year on advertising their products to consumers. Given the availability of consumer data in the internet, there is a trend to target campaigns to the most relevant online consumers so that company resources are conserved. Nevertheless, in offline media, this kind of individual targeting is not common, yet. Using real-world data from a company-run field experiment, we examine the size of selection bias and causal effects in a targeted offline advertising campaign. Furthermore, we analyze the return on advertising and compare the value of different underlying information for targeting. Similar to previous studies, we find considerable selection bias. Causal effects are in most - but not all - scenarios statistically significant. However, the analyzed campaign is not expected to be profitable because causal effects are too small. In contrast to other studies, the predictive value of information does not vary much in our results. This thesis highlights the importance of field experiments in advertising and shows the feasibility of offline targeting. |
|---|---|
| Autores principais: | Schaper, Katharina |
| Assunto: | Advertising Causal effect Field experiment Return on advertising Selection bias Targeting Television Value of information |
| Ano: | 2016 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso restrito |
| Instituição associada: | Universidade Católica Portuguesa |
| Idioma: | inglês |
| Origem: | Veritati - Repositório Institucional da Universidade Católica Portuguesa |
| Resumo: | Companies all over the world spend billions of dollars every year on advertising their products to consumers. Given the availability of consumer data in the internet, there is a trend to target campaigns to the most relevant online consumers so that company resources are conserved. Nevertheless, in offline media, this kind of individual targeting is not common, yet. Using real-world data from a company-run field experiment, we examine the size of selection bias and causal effects in a targeted offline advertising campaign. Furthermore, we analyze the return on advertising and compare the value of different underlying information for targeting. Similar to previous studies, we find considerable selection bias. Causal effects are in most - but not all - scenarios statistically significant. However, the analyzed campaign is not expected to be profitable because causal effects are too small. In contrast to other studies, the predictive value of information does not vary much in our results. This thesis highlights the importance of field experiments in advertising and shows the feasibility of offline targeting. |
|---|