SQL: a JOIN can multiply rows, and SUM will multiply with them

-- order.total is repeated once per item -> revenue is inflated
SELECT SUM(o.total)
  FROM orders o JOIN order_item i ON i.order_id = o.id;

-- aggregate each side separately
SELECT (SELECT SUM(total) FROM orders)      AS revenue,
       (SELECT COUNT(*)  FROM order_item)   AS items;

-- or de-duplicate before summing
SELECT SUM(total) FROM (
  SELECT DISTINCT o.id, o.total
    FROM orders o JOIN order_item i ON i.order_id = o.id) t;

joined SUM      : 4,182,110    <- wrong, counts each order once per item
separate SUM    :   482,110    <- correct

-- The tell is a revenue figure that is a suspiciously round multiple of the
-- real one - roughly the average number of items per order.