Publicação
Modeling patterns of subscription propensity of online newspapers: The case of Público newspaper
| Resumo: | This report describes developing a propensity model to subscribe to a Portuguese newspaper named Público. To develop it, the subscribers' behavior 30 days before subscribing was analyzed. Thus, Público could interact with the readers that present the same pattern as the ones that have already subscribed. The need for the development of these types of models is increasing, especially in the newspaper industry. The Digital Age has brought new challenges, one of which is adapting the subscription business model. This model forces newspapers to meet the needs of consumers by encouraging them to subscribe to an online newspaper. In this way, they are crucial for successfully implementing paywalls and increasing subscriptions. The chosen method to develop the model was Market Basket Analysis. This method was chosen due to its simplicity in adapting the business. The purpose was to find associations between the items that customers bought together. In this case, the aim is to find patterns in events performed together and see their association. After obtaining the patterns was conducted an A/B test with the control of two experimental groups. The experiment groups implemented two web pushes to incentive readers to subscribe. Although the probabilities of the Market Basket Analysis metrics are low due to the high number of different types of events, it was possible to find some interesting association rules. With these, an A/B test was conducted with a control group and two experimental groups, and it was concluded that one of these last two had an above average conversion rate. Thus, this project showed to be a possible and initial strategy to study and obtain patterns of customer behavior and mostly a possible strategy to overcome the challenges of the paywall implementation. |
|---|---|
| Autores principais: | Velho, Maria Inês Ramos |
| Assunto: | Subscribers Newspaper Propensity models Market Basket Analysis Events Conversion Rate |
| Ano: | 2023 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade Nova de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório Institucional da UNL |
| Resumo: | This report describes developing a propensity model to subscribe to a Portuguese newspaper named Público. To develop it, the subscribers' behavior 30 days before subscribing was analyzed. Thus, Público could interact with the readers that present the same pattern as the ones that have already subscribed. The need for the development of these types of models is increasing, especially in the newspaper industry. The Digital Age has brought new challenges, one of which is adapting the subscription business model. This model forces newspapers to meet the needs of consumers by encouraging them to subscribe to an online newspaper. In this way, they are crucial for successfully implementing paywalls and increasing subscriptions. The chosen method to develop the model was Market Basket Analysis. This method was chosen due to its simplicity in adapting the business. The purpose was to find associations between the items that customers bought together. In this case, the aim is to find patterns in events performed together and see their association. After obtaining the patterns was conducted an A/B test with the control of two experimental groups. The experiment groups implemented two web pushes to incentive readers to subscribe. Although the probabilities of the Market Basket Analysis metrics are low due to the high number of different types of events, it was possible to find some interesting association rules. With these, an A/B test was conducted with a control group and two experimental groups, and it was concluded that one of these last two had an above average conversion rate. Thus, this project showed to be a possible and initial strategy to study and obtain patterns of customer behavior and mostly a possible strategy to overcome the challenges of the paywall implementation. |
|---|