SQL Query to Find The First Order For All Customers

I have a table like this:

-order_id
-order_date
-user_id

For each user, I want to find their *first* order only.  order_id is an auto increment and order_date is datetime.  Lowest ID or earliest date would be their first order.
LVL 4
trippy1976Asked:
Who is Participating?

Improve company productivity with a Business Account.Sign Up

x
 
tlovieConnect With a Mentor Commented:
Okay, I think I understand what you want..

Step 1:

select user_id, min(order_date) as min_order_date from table group by user_id
-- this gives us the date that each customer orders a product

Step 2:

select count(user_id), MONTH(min_order_date), YEAR(min_order_date) from (
select user_id, min(order_date) as min_order_date from table group by user_id ) as q
group by MONTH(min_order_date), YEAR(min_order_date)
-- use the step 1 subquery to find all the new customers by month

Give that a go... all you need to execute is the step 2 query.
0
 
tlovieCommented:
SELECT user_id, min(order_date) as min_order_date
FROM table
GROUP BY user_id
0
 
tlovieCommented:
or you can do this:

SELECT user_id, min(order_id) as min_order_id
FROM table
GROUP BY user_id
0
The 14th Annual Expert Award Winners

The results are in! Meet the top members of our 2017 Expert Awards. Congratulations to all who qualified!

 
trippy1976Author Commented:
Excellent, thank you.  Before I close this can I impose for one more quickie?

I'd like to know how many customers made their first order in each month/year

Select count(min(order_id)) as total_orders, MONTH(order_date) as order_month, YEAR(order_date) as order_year
from table
group by user_id
order by order_year, order_month asc

??  Am I close?  

Thanks!
0
 
Guy Hengel [angelIII / a3]Billing EngineerCommented:
this article can help to write the query you want:http://www.experts-exchange.com/A_3203.html
0
 
tlovieCommented:
count, sum, avg, min, max are called aggregate functions.  you cannot have them in your group by clause.  MONTH, YEAR, etc, are scalar functions, so if they are in your select statement, they must also be in your group by clause.

I'm not sure if you still want user_id in there, but it could be in our out.... if you want to take it out, you would remove it from the SELECT and the GROUP BY clauses.

select user_id, count(order_id), MONTH(order_date) as order_month, YEAR(order_date) as order_year
from table
group by user_id, MONTH(order_date), YEAR(order_date)
order by order_year, order_month asc
0
 
trippy1976Author Commented:
Thanks for the great replies tlovie, but maybe I confused by trying it myself.  

What I want is only a count of the orders made in a month that are that customer's *first* order.

The first query shows each first order with the date, but I get something like :

1   9/1/2010
2  9/2/2010
3  9/3/2010

etc.

Where col 1 is the order_id and col 2 is the date of the order.  This is great.

Now I'd like something like:

3  9   2010

Where col1 is the total first time orders in month 9 (col2) of 2010 (col3)

Is this doable in a single query?
0
 
NerdsOfTechTechnology ScientistCommented:
Like tloive said if you want to NOT GROUP BY user_id just take it out of both the select and group by areas.

Also I think year should group out first - altered SQL below:
select count(order_id), MONTH(order_date) as order_month, YEAR(order_date) as order_year
from table
group YEAR(order_date), MONTH(order_date)
order by order_year, order_month asc

Open in new window

0
 
NerdsOfTechTechnology ScientistCommented:
however this give you ALL orders.

try something that will give you a count of the MINIMUMs

try this and see if we get any errors:
select user_id, YEAR(order_date) as order_year, MONTH(order_date) as order_month, SUM(IF(MIN(order_id) <> NULL, 1, 0)) as first_orders
from table
group by user_id, YEAR(order_date), MONTH(order_date), order_date
order by order_year, order_month asc

Open in new window

0
 
NerdsOfTechTechnology ScientistCommented:
oops i mean try:
select user_id, YEAR(order_date) as order_year, MONTH(order_date) as order_month, SUM(IF(MIN(order_id) <> NULL, 1, 0)) as first_orders
from table
group by user_id, YEAR(order_date), MONTH(order_date)
order by order_year, order_month asc

Open in new window

0
 
NerdsOfTechTechnology ScientistCommented:
then you can SUPER-query the results:

here is an example of trippy1976's query w/ SUPER-QUERY
SELECT SUM(first_orders), order_year, order_month FROM (
Select count(min(order_id)) as first_orders, MONTH(order_date) as order_month, YEAR(order_date) as order_year
from table
group by user_id, order_year, order_month
order by order_date asc
) x
group by order_year, order_month

Open in new window

0
 
NerdsOfTechTechnology ScientistCommented:
fixed
SELECT SUM(first_orders), order_year, order_month FROM (
Select count(min(order_id)) as first_orders, MONTH(order_date) as order_month, YEAR(order_date) as order_year
from table
group by user_id, order_year, order_month
order by order_year, order_month asc
) x
group by order_year, order_month

Open in new window

0
 
NerdsOfTechTechnology ScientistCommented:
nm order_date should sort
fixed.:
SELECT SUM(first_orders), order_year, order_month FROM (
Select count(min(order_id)) as first_orders, MONTH(order_date) as order_month, YEAR(order_date) as order_year
from table
group by user_id, order_year, order_month
order by order_date asc
) x
group by order_year, order_month

Open in new window

0
 
NerdsOfTechTechnology ScientistCommented:
fixed min parameter. try:
SELECT 
 SUM(first_orders) AS total_first_orders, 
 order_year, 
 order_month 
FROM 
(
 SELECT 
  COUNT(MIN(order_date)) AS first_orders, 
  MONTH(order_date) AS order_month, 
  YEAR(order_date) AS order_year
 FROM table
 GROUP BY user_id, YEAR(order_date), MONTH(order_date) 
 ORDER BY order_date ASC
) x
GROUP BY order_year ASC, order_month ASC

Open in new window

0
 
NerdsOfTechTechnology ScientistCommented:
the inner query selects first orders from EACH month from EACH customer.

say there is 3 customers:

first_order count|month|year

1|9|2010
1|10|2010
1|11|2010

1|9|2010
0|10|2010
1|11|2010

1|9|2010
0|10|2010
1|11|2010

Then the UPPER or SUPER query will take totals of matching year, month totals you end up with:

total first orders|year|month
3|2010|9
1|2010|10
3|2010|11
0
 
trippy1976Author Commented:
Perfect on both counts, thanks!
0
 
NerdsOfTechTechnology ScientistCommented:
Thanks for the feedback. Which query did you end up using in the end?
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.