Rather than having to retrain the model each time you have new data to test, a stored model allows you to load the trained model and make predictions without having to retrain the model