The most common solution is to frequently retrain on the latest data. A forecasting model might retrain every week, including the last weeks data, and might eve

The most common solution is to frequently retrain on the latest data. A forecasting model might retrain every week, including the last weeks data, and might even drop older data, for instance training data older than a year. It's best to transform your target variables, like "number of orders", to "number of orders per customer per day" or something like that. And then in your pipeline, you feed the latest estimate on your number of customers (e.g. average of the last two weeks). That's way more robust over time.