Tips on proving causation with customer engagement and satisfaction?

Looking at satisfaction surveys, we’ve seen that customers that are actively engaged with our social media and emails tend to have higher satisfaction scores. What are some ways I can determine causation? We have some control over future survey questions, or possibly follow up questions with existing survey takers. I’ve also considered looking for trends by topics they’ve engaged with. Any tips would be greatly appreciated!

causation customer engagement datascience tips