Question

I have the following query selecting everything from the notes table where the input(ex: bob) is not in the orders table.

SELECT * FROM `notes` WHERE notes.customer_email NOT IN 
(SELECT customers_email_address FROM orders) 
AND ((customer_phone LIKE '%bob%') 
OR (customer_name LIKE '%bob%') 
OR (customer_email LIKE '%bob%')) 
AND customers_id IS NULL 
GROUP BY `customer_email` 
ORDER BY `customer_name` 
DESC LIMIT 50

This fat boy of a query is taking ~80 seconds on my dev machine and ~7 seconds on the live server.

Two questions:

1. What did I do wrong with this query? (I need to learn from my mistakes)
2. How can I improve this query?

Answer 1

Try to use a join:

SELECT *
FROM
  `notes` left join orders
   on notes.customer_email=orders.customers_email_address
WHERE
  orders.customers_email_address is null
  AND notes.customers_id IS NULL 
  AND ((customer_phone LIKE '%bob%') 
        OR (customer_name LIKE '%bob%') 
        OR (customer_email LIKE '%bob%')) 
ORDER BY `customer_name` 
DESC LIMIT 50

they are usually faster than the IN/NOT IN clause. And also, I am not sure why you put a group by clause here.