mysql - Fixing SQL Query so it will become more Efficient -

i've got 3 tables:

1. mobile_users - id,phone_type,...

2+3. iphone_purchases , android_purchases - id,status,user_id,..

i trying of users made 2 or more purchases. successful purchase identified status > 0. tring total amount of users in mobile_users table in same query.

this query came with:

select count(*) `users`,        ( select count(*)             `mobile_users`        ) `total`   `mobile_users`   `mobile_users`.`phone_type` = 'iphone'    , ( select count(*)             ( select `status`,                          `user_id`                      `iphone_purchases`                   union                   select `status`,                          `user_id`                      `android_purchases`                 ) `purchase_list`            `purchase_list`.`status` > 0              , `purchase_list`.`user_id` = `mobile_users`.`id`        ) >= 2 

it's slow, , have find way improve it. appreciated!

edit: should take in consideration i'm building query sub-queries in php. i'm building more conditions on statment.

your query returning counts of users, not each user.

the following restructures query. counts number of purchases iphones , androids separately, , combines them using left outer join. where clause combines counts:

select mu.*, i.cnt iphones, a.cnt androids mobile_users mu left outer join      (select `user_id`, count(*) cnt       `iphone_purchases`       `status` > 0       group user_id      )      on i.user_id = mu.id left outer join      (select `user_id`, count(*) cnt       `android_purchases`       `status` > 0       group user_id      )      on a.user_id = mu.id coalesce(i.cnt, 0) + coalesce(a.cnt, 0) >= 2;