Joining orders to items creates several joined rows for an order with several items. Adding revenue over those rows would count the same order more than once. Looker uses primary keys and declared relationships for symmetric aggregates. Declare orders to items as one_to_many and keep unique keys. A relationship describes the actual data; choosing a label that makes one total pass does not repair a bad join.
Use the field picker and the generated result together. The warehouse contains repeated customer keys, orders with several items and missing values. Those rows make a plausible-looking total insufficient evidence. Run the requested definition, inspect its grain, and use the feedback to locate the mismatch. Every dataset here is fictional.
Keep order measures correct when an order has multiple items.