The rise of social media has transformed the way people communicate and share ideas. It has also changed how politicians interact with their followers and conduct propaganda to spread their messages. This has questioned social networks’ role in being a tool to not only measure how loved a candidate is but also whether they could be a viable option to predict electoral disputes. Various works have been performed in the field ever since Obama’s Twitter campaign in 2008, greatly dominated by sentiment analysis over a sample of text. A vast number of opposition has grown against this approach and those that derive from it. As a result, other methods have emerged. In the present work, we predict the 2024 Mexican Presidential election with both a regression algorithm and a neural network approach. Results show Claudia Sheinbaum will emerge as victorious with, in the closest scenario, 13 points above Xóchitl Gálvez. This is based on data 1 month before the election and experiments performed 3 days before election day.