I'm creating an average retention curve for a business. However right now, the denominator in every month is accounting for every customer that had a transaction in month 0 but the problem is after Month 0 not every customer has been a customer long enough to have a transaction in the subsequent months. How would I change this query so the denominator only accounts for customers to have been around long enough to have a transaction in that month? So for instance all customers that had their first transaction in July are removed from Month 1- Month 29 denominator?
Below is the query I'm currently using:
SELECT
cohorts.user_count
,CAST(cohorts.user_count AS DECIMAL(36,4)) / first_cohort.user_count AS retention_pct
,cohorts.transaction_count
,cohorts.Cohort
from
(
select
count(distinct t.owner) AS user_count
,count(t.id) AS transaction_count
,datediff(month,[f.first_transaction_date:month],[t.createdon:month]) as Cohort
from [transaction_cache as t]
JOIN
(
SELECT
owner
,MIN(createdon) AS first_transaction_date
FROM [transaction_cache]
WHERE [createdon:year] > ['2017-01-01':date:year]
GROUP BY 1
) AS f
ON f.owner = t.owner
where [t.createdon:year] > ['2017-01-01':date:year]
and t.status = 'successful'
and t.type = 'savings'
group by 3
) AS cohorts
JOIN
(
select
count(distinct t.owner) AS user_count
,count(t.id) AS transaction_count
,datediff(month,[f.first_transaction_date:month],[t.createdon:month]) as Cohort
from [transaction_cache as t]
JOIN
(
SELECT
owner
,MIN(createdon) AS first_transaction_date
FROM [transaction_cache]
WHERE [createdon:year] > ['2017-01-01':date:year]
GROUP BY 1
) AS f
ON f.owner = t.owner
where [t.createdon:year] > ['2017-01-01':date:year]
and t.status = 'successful'
and t.type = 'savings'
and datediff(month,[f.first_transaction_date:month],[t.createdon:month]) = 0
group by 3
) AS first_cohort
ON 1=1
order by 4 asc
Related
I'm facing an issue with the following query. It gave me this error [SELECT list expression references column integration_start_date which is neither grouped nor aggregated at [34:63]]. In particular, it points to the first 'when' in the result table, which I don't know how to fix. This is on BigQuery if that helps. I see everything is written correctly or I could be wrong. Seeking for help.
with plan_data as (
select format_date("%Y-%m-%d",last_day(date(a.basis_date))) as invoice_date,
a.sponsor_id as sponsor_id,
b.company_name as sponsor_name,
REPLACE(SUBSTR(d.meta,STRPOS(d.meta,'merchant_id')+12,13),'"','') as merchant_id,
a.state as plan_state,
date(c.start_date) as plan_start_date,
a.employee_id as square_employee_id,
date(
(select min(date)
from glproductionview.stats_sponsors
where sponsor_id = a.sponsor_id and sponsor_payroll_provider_identifier = 'square' and date >= c.start_date) )
as integration_start_date,
count(distinct a.employee_id) as eligible_pts_count, --pts that are in active plan and have payroll activities (payroll deductions) in the reporting month
from glproductionview.payroll_activities as a
left join glproductionview.sponsors as b
on a.sponsor_id = b.id
left join glproductionview.dc_plans as c
on a.plan_id = c.id
left join glproductionview.payroll_connections as d
on a.sponsor_id = d.sponsor_id and d.provider_identifier = 'rocket' and a.company_id = d.payroll_id
where a.payroll_provider_identifier = 'rocket'
and format_date("%Y-%m",date(a.basis_date)) = '2021-07'
and a.amount_cents > 0
group by 1,2,3,4,5,6,7,8
order by 2 asc
)
select invoice_date,
sponsor_id,
sponsor_name,
eligible_pts_count,
case
when eligible_pts_count <= 5 and date_diff(current_date(),integration_start_date, month) <= 12 then 20
when eligible_pts_count <= 5 and date_diff(current_date(),integration_start_date, month) > 12 then 15
when eligible_pts_count > 5 and date_diff(current_date(),integration_start_date, month) <= 12 then count(distinct square_employee_id)*4
when eligible_pts_count > 5 and date_diff(current_date(),integration_start_date, month) > 12 then count(distinct square_employee_id)*3
else 0
end as fees
from plan_data
group by 1,2,3,4;
I am doing a cohort analysis using the table TRANSACTIONS. Below is the table schema,
USER_ID NUMBER,
PAYMENT_DATE_UTC DATE,
IS_PAYMENT_ADDED BOOLEAN
Below is a quick query to see how USER_ID 12345 (an example) goes through the different cohorts based on the date filter provided,
WITH RESULT(
SELECT
USER_ID,
TO_DATE(PAYMENT_DATE_UTC) AS PAYMENT_DATE,
SUM(CASE WHEN IS_PAYMENT_ADDED=TRUE THEN 1 ELSE 0 END) AS PAYMENT_ADDED_COUNT
FROM TRANSACTIONS
GROUP BY 1,2
HAVING PAYMENT_ADDED_COUNT>=1
ORDER BY 2
)
SELECT
COUNT(DISTINCT r.USER_ID),
SUM(r.PAYMENT_ADDED_COUNT)
FROM RESULT r
WHERE r.USER_ID=12345
AND (r.PAYMENT_DATE>='2021-02-01' AND r.PAYMENT_DATE<'2021-02-15')
The result for this query with the time frame (two weeks) would be
| 1 | 55 |
and this USER_ID would be classified as a Regular User Cohort (one who has made more than 10 payments) for the provided date filter
If the same query is run with the time frame as just one day say '2021-02-07', the result would be
| 1 | 10 |
and this USER_ID would be classified as as Occasional User Cohort (one who has made between 1 and 10 payments) for the provided date filter
I have this below query to bucket the USER_ID's into the two different cohorts based on the sum of the payments added,
WITH
ALL_USER_COHORT AS
(SELECT
USER_ID,
SUM(CASE WHEN IS_PAYMENT_ADDED=TRUE THEN 1 ELSE 0 END ) AS PAYMENT_ADDED_COUNT
FROM TRANSACTIONS
GROUP BY USER_ID
),
OCASSIONAL_USER_COHORT AS
(SELECT
USER_ID,
SUM(CASE WHEN IS_PAYMENT_ADDED=TRUE THEN 1 ELSE 0 END ) AS PAYMENT_ADDED_COUNT
FROM TRANSACTIONS
GROUP BY USER_ID
HAVING (PAYMENT_ADDED_COUNT>=1 AND PAYMENT_ADDED_COUNT<=10)
),
REGULAR_USER_COHORT AS
(SELECT
USER_ID,
SUM(CASE WHEN IS_PAYMENT_ADDED=TRUE THEN 1 ELSE 0 END ) AS PAYMENT_ADDED_COUNT
FROM TRANSACTIONS
GROUP BY USER_ID
HAVING PAYMENT_ADDED_COUNT>10
)
SELECT
COUNT(DISTINCT ou.USER_ID) AS "OCCASIONAL USERS",
COUNT(DISTINCT ru.USER_ID) AS "REGULAR USERS"
FROM ALL_USER_COHORT au
LEFT JOIN OCASSIONAL_USER_COHORT ou ON au.USER_ID=ou.USER_ID
LEFT JOIN REGULAR_USER_COHORT ru ON au.USER_ID=ru.USER_ID
LEFT JOIN TRANSACTIONS t ON au.USER_ID=t.USER_ID
WHERE au.USER_ID=12345
AND TO_DATE(t.PAYMENT_DATE_UTC)>='2021-02-07'
Ideally the USER_ID 12345 should be bucketed as "OCCASIONAL USERS" as per the provided date filter but the query buckets it as "REGULAR USERS" instead.
For starters you CTE could have the redundancy removed like so:
WITH all_user_cohort AS (
SELECT
USER_ID,
SUM(IFF(is_payment_added=TRUE, 1,0)) AS payment_added_count
FROM transactions
GROUP BY user_id
), ocassional_user_cohort AS (
SELECT * FROM all_user_cohort
WHERE PAYMENT_ADDED_COUNT between 1 AND 10
), regular_user_cohort AS (
SELECT * FROM all_user_cohort
WHERE PAYMENT_ADDED_COUNT > 10
)
SELECT
COUNT(DISTINCT ou.user_id) AS "OCCASIONAL USERS",
COUNT(DISTINCT ru.user_id) AS "REGULAR USERS"
FROM all_user_cohort AS au
LEFT JOIN ocassional_user_cohort ou ON au.user_id=ou.user_id
LEFT JOIN regular_user_cohort ru ON au.user_id=ru.user_id
LEFT JOIN transactions t ON au.user_id=t.user_id
WHERE au.user_id=12345
AND TO_DATE(t.payment_date_utc)>='2021-03-01'
But the reason you are getting this problem is you are doing the which do the belong in across all time.
What you are wanting is to move the date filter into all_user_cohort, and not making tables when you can just sum the number of rows meeting the need.
WITH all_user_cohort AS (
SELECT
USER_ID,
SUM(IFF(is_payment_added=TRUE, 1,0)) AS payment_added_count
FROM transactions
WHERE TO_DATE(payment_date_utc)>='2021-03-01'
GROUP BY user_id
)
SELECT
SUM(IFF(payment_added_count between 1 AND 10, 1,0)) AS "OCCASIONAL USERS"
SUM(IFF(payment_added_count > 10, 1,0)) AS "REGULAR USERS"
FROM transactions
WHERE au.user_id=12345
Which can also be done differently, if that is more what your looking for, for other reasons.
WITH all_user_cohort AS (
SELECT
USER_ID,
SUM(IFF(is_payment_added=TRUE, 1,0)) AS payment_added_count
FROM transactions
WHERE TO_DATE(payment_date_utc)>='2021-03-01'
GROUP BY user_id
), classify_users AS (
SELECT user_id
,CASE
WHEN payment_added_count between 1 AND 10 THEN 'OCCASIONAL USERS'
WHEN payment_added_count > 10 THEN 'REGULAR USERS'
ELSE 'users with zero payments'
END AS classified
FROM all_user_cohort
)
SELECT classified
,count(*)
FROM classify_users
WHERE user_id=12345
GROUP BY 1
I'm trying to use a nested aggregate function. I know that SQL does not support it, but I really need to do something like the below query. Basically, I want to count the number of users for each day. But I want to only count the users that haven't completed an order within a 15 days window (relative to a specific day) and that have completed any order within a 30 days window (relative to a specific day). I already know that it is not possible to solve this problem using a regular subquery (it does not allow to change subquery values for each date). The "id" and the "state" attributes are related to the orders. Also, I'm using Fivetran with Snowflake.
SELECT
db.created_at::date as Date,
count(case when
(count(case when (db.state = 'finished')
and (db.created_at::date between dateadd(day,-15,Date) and dateadd(day,-1,Date)) then db.id end)
= 0) and
(count(case when (db.state = 'finished')
and (db.created_at::date between dateadd(day,-30,Date) and dateadd(day,-16,Date)) then db.id end)
> 0) then db.user end)
FROM
data_base as db
WHERE
db.created_at::date between '2020-01-01' and dateadd(day,-1,current_date)
GROUP BY Date
In other words, I want to transform the below query in a way that the "current_date" changes for each date.
WITH completed_15_days_before AS (
select
db.user as User,
count(case when db.state = 'finished' then db.id end) as Completed
from
data_base as db
where
db.created_at::date between dateadd(day,-15,current_date) and dateadd(day,-1,current_date)
group by User
),
completed_16_days_before AS (
select
db.user as User,
count(case when db.state = 'finished' then db.id end) as Completed
from
data_base as db
where
db.created_at::date between dateadd(day,-30,current_date) and dateadd(day,-16,current_date)
group by User
)
SELECT
date(db.created_at) as Date,
count(distinct case when comp_15.completadas = 0 and comp_16.completadas > 0 then comp_15.user end) as "Total Users Churn",
count(distinct case when comp_15.completadas > 0 then comp_15.user end) as "Total Users Active",
week(Date) as Week
FROM
data_base as db
left join completadas_15_days_before as comp_15 on comp_15.user = db.user
left join completadas_16_days_before as comp_16 on comp_16.user = db.user
WHERE
db.created_at::date between '2020-01-01' and dateadd(day,-1,current_date)
GROUP BY Date
Does anyone have a clue on how to solve this puzzle? Thank you very much!
The following should give you roughly what you want - difficult to test without sample data but should be a good enough starting point for you to then amend it to give you exactly what you want.
I've commented to the code to hopefully explain what each section is doing.
-- set parameter for the first date you want to generate the resultset for
set start_date = TO_DATE('2020-01-01','YYYY-MM-DD');
-- calculate the number of days between the start_date and the current date
set num_days = (Select datediff(day, $start_date , current_date()+1));
--generate a list of all the dates from the start date to the current date
-- i.e. every date that needs to appear in the resultset
WITH date_list as (
select
dateadd(
day,
'-' || row_number() over (order by null),
dateadd(day, '+1', current_date())
) as date_item
from table (generator(rowcount => ($num_days)))
)
--Create a list of all the orders that are in scope
-- i.e. 30 days before the start_date up to the current date
-- amend WHERE clause to in/exclude records as appropriate
,order_list as (
SELECT created_at, rt_id
from data_base
where created_at between dateadd(day,-30,$start_date) and current_date()
and state = 'finished'
)
SELECT dl.date_item
,COUNT (DISTINCT ol30.RT_ID) AS USER_COUNT
,COUNT (ol30.RT_ID) as ORDER_COUNT
FROM date_list dl
-- get all orders between -30 and -16 days of each date in date_list
left outer join order_list ol30 on ol30.created_at between dateadd(day,-30,dl.date_item) and dateadd(day,-16,dl.date_item)
-- exclude records that have the same RT_ID as in the ol30 dataset but have a date between 0 amd -15 of the date in date_list
WHERE NOT EXISTS (SELECT ol15.RT_ID
FROM order_list ol15
WHERE ol30.RT_ID = ol15.RT_ID
AND ol15.created_at between dateadd(day,-15,dl.date_item) and dl.date_item)
GROUP BY dl.date_item
ORDER BY dl.date_item;
I want to retrieve records where cash deposits are more than 4 totaling to 1000000 during a day and continues for more than 5 days.
I have came up with below query.
SELECT COUNT(a.txamt) AS "txcount"
, SUM(a.txamt) AS "txsum"
, b.custcd
, a.txdate
FROM tb_transactions a
INNER JOIN tb_accounts b
ON a.acctno = b.acctno
WHERE a.cashflowtype = 'CR'
GROUP BY b.custcd, a.txdate
HAVING COUNT(a.txamt)>4 and SUM(a.txamt)>='1000000'
ORDER BY a.txdate;
But I'm stuck on how to fetch the records if the pattern continues for 5 days.
How to achieve the desired result?
Something like:
SELECT *
FROM (
SELECT t.*,
COUNT( txdate ) OVER ( PARTITION BY custcd
ORDER BY txdate
RANGE BETWEEN INTERVAL '0' DAY PRECEDING
AND INTERVAL '4' DAY FOLLOWING ) AS
num_days
FROM (
select count(a.txamt) as "txcount",
sum(a.txamt) as "txsum",
b.custcd,
a.txdate
from tb_transactions a inner join tb_accounts b on a.acctno=b.acctno
where a.cashflowtype='CR'
group by b.custcd, a.txdate
having count(a.txamt)>4 and sum(a.txamt)>=1000000
) t
)
WHERE num_days = 5
order by a.txdate;
A "Case" in CRM has a field called "Status" with four options.
I'm trying to
build a report in CRM that fills a table with every week of the year (each row is a different week), and then counts the number of cases that have each Status option (the columns would be each of the Status options).
The table would look like this
Status 1 Status 2 Status 3
Week 1 3 55 4
Week 2 5 23 5
Week 3 14 11 33
So far I have the following:
SELECT
SUM(case WHEN status = 1 then 1 else 0 end) Status1,
SUM(case WHEN status = 2 then 1 else 0 end) Status2,
SUM(case WHEN status = 3 then 1 else 0 end) Status3,
SUM(case WHEN status = 4 then 1 else 0 end) Status4,
SUM(case WHEN status = 5 then 1 else 0 end) Status5
FROM [DB].[dbo].[Contact]
Which gives me the following:
Status 1 Status 2 Status 3
2 43 53
Now I need to somehow split this into 52 rows for the past year and filter these results by date (columns in the Contact table). I'm a bit new to SQL queries and CRM - any help here would be much appreciated.
Here is a SQLFiddle with my progress and sample data: http://sqlfiddle.com/#!2/85b19/1
Sounds like you want to group by a range. The trick is to create a new field that represents each range (for you one per year) and group by that.
Since it also seems like you want an infinite range of dates, marc_s has a good summary for how to do the group by trick with dates in a generic way: SQL group by frequency within a date range
So, let's break this down:
You want to make a report that shows, for each contact, a breakdown, week by week, of the number of cases registered to that contact, which is divided into three columns, one for each StateCode.
If this is the case, then you would need to have 52 date records (or so) for each contact. For calendar like requests, it's always good to have a separate calendar table that lets you query from it. Dan Guzman has a blog entry that creates a useful calendar table which I'll use in the query.
WITH WeekNumbers AS
(
SELECT
FirstDateOfWeek,
-- order by first date of week, grouping calendar year to produce week numbers
WeekNumber = row_number() OVER (PARTITION BY CalendarYear ORDER BY FirstDateOfWeek)
FROM
master.dbo.Calendar -- created from script
GROUP BY
FirstDateOfWeek,
CalendarYear
), Calendar AS
(
SELECT
WeekNumber =
(
SELECT
WeekNumber
FROM
WeekNumbers WN
WHERE
C.FirstDateOfWeek = WN.FirstDateOfWeek
),
*
FROM
master.dbo.Calendar C
WHERE
CalendarDate BETWEEN '1/1/2012' AND getutcdate()
)
SELECT
C.FullName,
----include the below if the data is necessary
--Cl.WeekNumber,
--Cl.CalendarYear,
--Cl.FirstDateOfWeek,
--Cl.LastDateOfWeek,
'Week: ' + CAST(Cl.WeekNumber AS VARCHAR(20))
+ ', Year: ' + CAST(Cl.CalendarYear AS VARCHAR(20)) WeekNumber
FROM
CRM.dbo.Contact C
-- use a cartesian join to produce a table list
CROSS JOIN
(
SELECT
DISTINCT WeekNumber,
CalendarYear,
FirstDateOfWeek,
LastDateOfWeek
FROM
Calendar
) Cl
ORDER BY
C.FullName,
Cl.WeekNumber
This is different from the solution Ben linked to because Marc's query only returns weeks where there is a matching value, whereas you may or may not want to see even the weeks where there is no activity.
Once you have your core tables of contacts split out week by week as in the above (or altered for your specific time period), you can simply add a subquery for each StateCode to see the breakdown in columns as in the final query below.
WITH WeekNumbers AS
(
SELECT
FirstDateOfWeek,
WeekNumber = row_number() OVER (PARTITION BY CalendarYear ORDER BY FirstDateOfWeek)
FROM
master.dbo.Calendar
GROUP BY
FirstDateOfWeek,
CalendarYear
), Calendar AS
(
SELECT
WeekNumber =
(
SELECT
WeekNumber
FROM
WeekNumbers WN
WHERE
C.FirstDateOfWeek = WN.FirstDateOfWeek
),
*
FROM
master.dbo.Calendar C
WHERE
CalendarDate BETWEEN '1/1/2012' AND getutcdate()
)
SELECT
C.FullName,
--Cl.WeekNumber,
--Cl.CalendarYear,
--Cl.FirstDateOfWeek,
--Cl.LastDateOfWeek,
'Week: ' + CAST(Cl.WeekNumber AS VARCHAR(20)) +', Year: ' + CAST(Cl.CalendarYear AS VARCHAR(20)) WeekNumber,
(
SELECT
count(*)
FROM
CRM.dbo.Incident I
INNER JOIN CRM.dbo.StringMap SM ON
I.StateCode = SM.AttributeValue
INNER JOIN
(
SELECT
DISTINCT ME.Name,
ME.ObjectTypeCode
FROM
CRM.MetadataSchema.Entity ME
) E ON
SM.ObjectTypeCode = E.ObjectTypeCode
WHERE
I.ModifiedOn >= Cl.FirstDateOfWeek
AND I.ModifiedOn < dateadd(day, 1, Cl.LastDateOfWeek)
AND E.Name = 'incident'
AND SM.AttributeName = 'statecode'
AND SM.LangId = 1033
AND I.CustomerId = C.ContactId
AND SM.Value = 'Active'
) ActiveCases,
(
SELECT
count(*)
FROM
CRM.dbo.Incident I
INNER JOIN CRM.dbo.StringMap SM ON
I.StateCode = SM.AttributeValue
INNER JOIN
(
SELECT
DISTINCT ME.Name,
ME.ObjectTypeCode
FROM
CRM.MetadataSchema.Entity ME
) E ON
SM.ObjectTypeCode = E.ObjectTypeCode
WHERE
I.ModifiedOn >= Cl.FirstDateOfWeek
AND I.ModifiedOn < dateadd(day, 1, Cl.LastDateOfWeek)
AND E.Name = 'incident'
AND SM.AttributeName = 'statecode'
AND SM.LangId = 1033
AND I.CustomerId = C.ContactId
AND SM.Value = 'Resolved'
) ResolvedCases,
(
SELECT
count(*)
FROM
CRM.dbo.Incident I
INNER JOIN CRM.dbo.StringMap SM ON
I.StateCode = SM.AttributeValue
INNER JOIN
(
SELECT
DISTINCT ME.Name,
ME.ObjectTypeCode
FROM
CRM.MetadataSchema.Entity ME
) E ON
SM.ObjectTypeCode = E.ObjectTypeCode
WHERE
I.ModifiedOn >= Cl.FirstDateOfWeek
AND I.ModifiedOn < dateadd(day, 1, Cl.LastDateOfWeek)
AND E.Name = 'incident'
AND SM.AttributeName = 'statecode'
AND SM.LangId = 1033
AND I.CustomerId = C.ContactId
AND SM.Value = 'Canceled'
) CancelledCases
FROM
CRM.dbo.Contact C
CROSS JOIN
(
SELECT
DISTINCT WeekNumber,
CalendarYear,
FirstDateOfWeek,
LastDateOfWeek
FROM
Calendar
) Cl
ORDER BY
C.FullName,
Cl.WeekNumber