Twitter as a marketing tool has led to a growing interest in measuring the effectiveness of content marketing on this platform. However, there has yet to be a comprehensive analytical model to measure the effectiveness of public content marketing (PCM) accurately and reliably. A literature review determined the gaps between preliminary studies and constructing a new model to measure the content effectiveness, considering variables related to interactivity and performance of digital content marketing (DCM) strategies. For this reason, this study aims to build an analytical model that determines which content characteristics improve the effectiveness of Twitter accounts, taking as a case study the governorates of Colombia. Within the methodology for data mining, CRISP-DM was used, which allowed the cleaning, processing and analysis of all data collected from the accounts of Colombian governments. The results allowed to establish factors that have yet to be considered to measure the Engagement Rate per Post (ERP) and have a critical load on users’ interactivity with the content, such as the tweet type, emojis, dates, the type of media, sentiment associated with the post and emotions. With the model, it was possible to identify the variables that improve the ERP and their impact on the effectiveness of the content.

Twitter 作为一种营销工具，引起了人们对衡量该平台上内容营销有效性的兴趣日益增长。然而，目前还没有一个全面的分析模型来准确可靠地衡量公共内容营销（PCM）的有效性。文献综述确定了初步研究与构建衡量内容有效性的新模型之间的差距，并考虑了与数字内容营销 (DCM) 策略的交互性和绩效相关的变量。因此，本研究旨在建立一个分析模型，以哥伦比亚各省为例，确定哪些内容特征可以提高 Twitter 帐户的有效性。在数据挖掘方法中，使用了 CRISP-DM，它可以进行清理、处理和分析从哥伦比亚政府账户收集的所有数据。结果允许建立尚未考虑的因素来衡量每个帖子的参与率 (ERP)，并对用户与内容的交互性产生关键负载，例如推文类型、表情符号、日期、媒体类型、与帖子和情绪相关的情绪。通过该模型，可以识别改进 ERP 的变量及其对内容有效性的影响。