The term "latent variables" is primarily used in the fields of artificial intelligence, big data, smart data and digital transformation. Latent variables are characteristics or influences that we cannot measure directly, but which are nevertheless important for our analyses.
Imagine you want to know how satisfied people are with an online shop. Satisfaction itself is not directly measurable, but it influences measurable things such as click behaviour, how long users stay on a page and how often they buy something. We can record this measurable data, but the underlying satisfaction remains "latent", i.e. hidden.
Latent variables are used in artificial intelligence to recognise correlations that are not visible at first glance. With big data in particular, they help to create an overall picture from a large amount of individual information, which supports companies in making decisions. Latent variables therefore make it possible to gain important insights from data, even if not all the information is obviously available.