주요 Google BigQuery 기능 개요 — 마케팅 분석을 위한 요청 작성 연습

기술 교육이나 SQL 지식 없이 GBQ 데이터에 대한 보고서를 작성합니다.

정기적으로 광고 캠페인에 대한 보고서가 필요하지만 SQL을 연구하거나 분석가의 응답을 기다릴 시간이 없으십니까? OWOX BI를 사용하면 데이터 구조를 이해할 필요 없이 보고서를 작성할 수 있습니다. 스마트 데이터 보고서에서 보고 싶은 매개변수와 측정항목을 선택하기만 하면 됩니다. OWOX BI Smart Data는 당신이 이해할 수 있는 방식으로 데이터를 즉시 시각화합니다.

OWOX BI를 무료로 사용해 보세요

목차

• SQL이란 무엇이며 BigQuery가 지원하는 방언
• 어디서 시작하나요
• Google BigQuery 기능
• 집계 함수
• 날짜 함수
• 문자열 함수
• 창 기능
• 결론

SQL이란 무엇이며 BigQuery가 지원하는 방언

어디서 시작하나요

독자를 위한 보너스

SQL 쿼리 실습을 위한 데모 데이터

지금 다운로드

OWOX BI는 다양한 시스템의 데이터를 BigQuery로 결합하는 데 도움이 됩니다. 웹사이트의 사용자 작업, 통화, CRM의 주문, 이메일, 광고 비용에 대한 데이터입니다. OWOX BI를 사용하여 고급 분석을 사용자 정의하고 복잡한 보고서를 자동화할 수 있습니다.

데모 받기

Google BigQuery 기능

집계 함수

    SELECT AVG(revenue) as average_revenue, MAX(revenue) as max_revenue, MIN(revenue) as min_revenue, SUM(revenue) as whole_revenue, COUNT(transactionId) as transactions, EXACT_COUNT_DISTINCT(transactionId) as unique_transactions FROM [owox-analytics:t_kravchenko.Demo_data]
SELECT AVG(revenue) as average_revenue, MAX(revenue) as max_revenue, MIN(revenue) as min_revenue, SUM(revenue) as whole_revenue, COUNT(transactionId) as transactions, EXACT_COUNT_DISTINCT(transactionId) as unique_transactions FROM [owox-analytics:t_kravchenko.Demo_data]

    SELECT AVG(revenue) as average_revenue, MAX(revenue) as max_revenue, MIN(revenue) as min_revenue, SUM(revenue) as whole_revenue, COUNT(transactionId) as transactions, COUNT(DISTINCT(transactionId)) as unique_transactions FROM `owox-analytics.t_kravchenko.Demo_data`
SELECT AVG(revenue) as average_revenue, MAX(revenue) as max_revenue, MIN(revenue) as min_revenue, SUM(revenue) as whole_revenue, COUNT(transactionId) as transactions, COUNT(DISTINCT(transactionId)) as unique_transactions FROM `owox-analytics.t_kravchenko.Demo_data`

독자를 위한 보너스

SQL 쿼리 실습을 위한 데모 데이터

지금 다운로드

날짜 함수

    SELECT CURRENT_DATE() AS today, DATE( date_UTC ) AS date_UTC_in_YYYYMMDD, DATE_ADD( date_UTC,1, 'DAY') AS date_UTC_plus_one_day, DATE_ADD( date_UTC,-1, 'DAY') AS date_UTC_minus_one_day, DATEDIFF(CURRENT_DATE(), date_UTC ) AS difference_between_date, DAY( CURRENT_DATE() ) AS the_day, MONTH( CURRENT_DATE()) AS the_month, YEAR( CURRENT_DATE()) AS the_year FROM [owox-analytics:t_kravchenko.Demo_data]
SELECT CURRENT_DATE() AS today, DATE( date_UTC ) AS date_UTC_in_YYYYMMDD, DATE_ADD( date_UTC,1, 'DAY') AS date_UTC_plus_one_day, DATE_ADD( date_UTC,-1, 'DAY') AS date_UTC_minus_one_day, DATEDIFF(CURRENT_DATE(), date_UTC ) AS difference_between_date, DAY( CURRENT_DATE() ) AS the_day, MONTH( CURRENT_DATE()) AS the_month, YEAR( CURRENT_DATE()) AS the_year FROM [owox-analytics:t_kravchenko.Demo_data]

    SELECT today, date_UTC_in_YYYYMMDD, DATE_ADD( date_UTC_in_YYYYMMDD, INTERVAL 1 DAY) AS date_UTC_plus_one_day, DATE_SUB( date_UTC_in_YYYYMMDD,INTERVAL 1 DAY) AS date_UTC_minus_one_day, DATE_DIFF(today, date_UTC_in_YYYYMMDD, DAY) AS difference_between_date, EXTRACT(DAY FROM today ) AS the_day, EXTRACT(MONTH FROM today ) AS the_month, EXTRACT(YEAR FROM today ) AS the_year FROM ( SELECT CURRENT_DATE() AS today, DATE( date_UTC ) AS date_UTC_in_YYYYMMDD FROM `owox-analytics.t_kravchenko.Demo_data`)
SELECT today, date_UTC_in_YYYYMMDD, DATE_ADD( date_UTC_in_YYYYMMDD, INTERVAL 1 DAY) AS date_UTC_plus_one_day, DATE_SUB( date_UTC_in_YYYYMMDD,INTERVAL 1 DAY) AS date_UTC_minus_one_day, DATE_DIFF(today, date_UTC_in_YYYYMMDD, DAY) AS difference_between_date, EXTRACT(DAY FROM today ) AS the_day, EXTRACT(MONTH FROM today ) AS the_month, EXTRACT(YEAR FROM today ) AS the_year FROM ( SELECT CURRENT_DATE() AS today, DATE( date_UTC ) AS date_UTC_in_YYYYMMDD FROM `owox-analytics.t_kravchenko.Demo_data`)

문자열 함수

    SELECT CONCAT(the_day,'-',the_month,'-',the_year) AS mix_string1 FROM ( SELECT 31 AS the_day, 12 AS the_month, 2018 AS the_year FROM [owox-analytics:t_kravchenko.Demo_data]) GROUP BY mix_string1
SELECT CONCAT(the_day,'-',the_month,'-',the_year) AS mix_string1 FROM ( SELECT 31 AS the_day, 12 AS the_month, 2018 AS the_year FROM [owox-analytics:t_kravchenko.Demo_data]) GROUP BY mix_string1

    SELECT CONCAT(the_day,'-',the_month,'-',the_year) AS mix_string1 FROM ( SELECT 31 AS the_day, 12 AS the_month, 2018 AS the_year FROM owox-analytics.t_kravchenko.Demo_data) GROUP BY mix_string1
SELECT CONCAT(the_day,'-',the_month,'-',the_year) AS mix_string1 FROM ( SELECT 31 AS the_day, 12 AS the_month, 2018 AS the_year FROM owox-analytics.t_kravchenko.Demo_data) GROUP BY mix_string1

    SELECT COUNT(transactionId) AS transactions, check FROM ( SELECT transactionId, page CONTAINS 'shop_id' AS check FROM [owox-analytics:t_kravchenko.Demo_data]) GROUP BY check
SELECT COUNT(transactionId) AS transactions, check FROM ( SELECT transactionId, page CONTAINS 'shop_id' AS check FROM [owox-analytics:t_kravchenko.Demo_data]) GROUP BY check

    SELECT COUNT(transactionId) AS transactions, check1, check2 FROM ( SELECT transactionId, REGEXP_CONTAINS( page, 'shop_id') AS check1, page LIKE '%shop_id%' AS check2 FROM `owox-analytics.t_kravchenko.Demo_data`) GROUP BY check1, check2
SELECT COUNT(transactionId) AS transactions, check1, check2 FROM ( SELECT transactionId, REGEXP_CONTAINS( page, 'shop_id') AS check1, page LIKE '%shop_id%' AS check2 FROM `owox-analytics.t_kravchenko.Demo_data`) GROUP BY check1, check2

예 2 . 각 전달 방법에 delivery_id를 할당했으며 페이지 URL에서 이 매개변수의 값을 지정합니다. 사용자가 선택한 배송 방법을 확인하려면 별도의 열에서 delivery_id를 선택해야 합니다.

이를 위해 다음 쿼리를 사용할 수 있습니다.

#기존 SQL

    SELECT page_lower_case, page_length, index_of_delivery_id, selected_delivery_id, REPLACE(selected_delivery_id, 'selected_delivery_id=', '') as delivery_id FROM ( SELECT page_lower_case, page_length, index_of_delivery_id, SUBSTR(page_lower_case, index_of_delivery_id) AS selected_delivery_id FROM ( SELECT page_lower_case, LENGTH(page_lower_case) AS page_length, INSTR(page_lower_case, 'selected_delivery_id') AS index_of_delivery_id FROM ( SELECT LOWER( page) AS page_lower_case, UPPER( page) AS page_upper_case FROM [owox-analytics:t_kravchenko.Demo_data]))) ORDER BY page_lower_case ASC
SELECT page_lower_case, page_length, index_of_delivery_id, selected_delivery_id, REPLACE(selected_delivery_id, 'selected_delivery_id=', '') as delivery_id FROM ( SELECT page_lower_case, page_length, index_of_delivery_id, SUBSTR(page_lower_case, index_of_delivery_id) AS selected_delivery_id FROM ( SELECT page_lower_case, LENGTH(page_lower_case) AS page_length, INSTR(page_lower_case, 'selected_delivery_id') AS index_of_delivery_id FROM ( SELECT LOWER( page) AS page_lower_case, UPPER( page) AS page_upper_case FROM [owox-analytics:t_kravchenko.Demo_data]))) ORDER BY page_lower_case ASC

    SELECT page_lower_case, page_length, index_of_delivery_id, selected_delivery_id, REPLACE(selected_delivery_id, 'selected_delivery_id=', '') AS delivery_id FROM ( SELECT page_lower_case, page_length, index_of_delivery_id, SUBSTR(page_lower_case, index_of_delivery_id) AS selected_delivery_id FROM ( SELECT page_lower_case, CHAR_LENGTH(page_lower_case) AS page_length, STRPOS(page_lower_case, 'selected_delivery_id') AS index_of_delivery_id FROM ( SELECT LOWER( page) AS page_lower_case, UPPER( page) AS page_upper_case FROM `owox-analytics.t_kravchenko.Demo_data`))) ORDER BY page_lower_case ASC
SELECT page_lower_case, page_length, index_of_delivery_id, selected_delivery_id, REPLACE(selected_delivery_id, 'selected_delivery_id=', '') AS delivery_id FROM ( SELECT page_lower_case, page_length, index_of_delivery_id, SUBSTR(page_lower_case, index_of_delivery_id) AS selected_delivery_id FROM ( SELECT page_lower_case, CHAR_LENGTH(page_lower_case) AS page_length, STRPOS(page_lower_case, 'selected_delivery_id') AS index_of_delivery_id FROM ( SELECT LOWER( page) AS page_lower_case, UPPER( page) AS page_upper_case FROM `owox-analytics.t_kravchenko.Demo_data`))) ORDER BY page_lower_case ASC

독자를 위한 보너스

SQL 쿼리 실습을 위한 데모 데이터

지금 다운로드

창 기능

    SELECT date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions FROM ( SELECT date, clientId, DayTime, AVG(revenue) OVER (PARTITION BY date, clientId, DayTime) AS avg_revenue, MAX(revenue) OVER (PARTITION BY date, clientId, DayTime) AS max_revenue, MIN(revenue) OVER (PARTITION BY date, clientId, DayTime) AS min_revenue, SUM(revenue) OVER (PARTITION BY date, clientId, DayTime) AS sum_revenue, COUNT(transactionId) OVER (PARTITION BY date, clientId, DayTime) AS transactions, COUNT(DISTINCT(transactionId)) OVER (PARTITION BY date, clientId, DayTime) AS unique_transactions FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM [owox-analytics:t_kravchenko.Demo_data])) GROUP BY date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions ORDER BY transactions DESC
SELECT date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions FROM ( SELECT date, clientId, DayTime, AVG(revenue) OVER (PARTITION BY date, clientId, DayTime) AS avg_revenue, MAX(revenue) OVER (PARTITION BY date, clientId, DayTime) AS max_revenue, MIN(revenue) OVER (PARTITION BY date, clientId, DayTime) AS min_revenue, SUM(revenue) OVER (PARTITION BY date, clientId, DayTime) AS sum_revenue, COUNT(transactionId) OVER (PARTITION BY date, clientId, DayTime) AS transactions, COUNT(DISTINCT(transactionId)) OVER (PARTITION BY date, clientId, DayTime) AS unique_transactions FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM [owox-analytics:t_kravchenko.Demo_data])) GROUP BY date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions ORDER BY transactions DESC

    #standardSQL SELECT date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions FROM ( SELECT date, clientId, DayTime, AVG(revenue) OVER (PARTITION BY date, clientId, DayTime) AS avg_revenue, MAX(revenue) OVER (PARTITION BY date, clientId, DayTime) AS max_revenue, MIN(revenue) OVER (PARTITION BY date, clientId, DayTime) AS min_revenue, SUM(revenue) OVER (PARTITION BY date, clientId, DayTime) AS sum_revenue, COUNT(transactionId) OVER (PARTITION BY date, clientId, DayTime) AS transactions, COUNT(DISTINCT(transactionId)) OVER (PARTITION BY date, clientId, DayTime) AS unique_transactions FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM `owox-analytics.t_kravchenko.Demo_data`)) GROUP BY date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions ORDER BY transactions DESC
#standardSQL SELECT date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions FROM ( SELECT date, clientId, DayTime, AVG(revenue) OVER (PARTITION BY date, clientId, DayTime) AS avg_revenue, MAX(revenue) OVER (PARTITION BY date, clientId, DayTime) AS max_revenue, MIN(revenue) OVER (PARTITION BY date, clientId, DayTime) AS min_revenue, SUM(revenue) OVER (PARTITION BY date, clientId, DayTime) AS sum_revenue, COUNT(transactionId) OVER (PARTITION BY date, clientId, DayTime) AS transactions, COUNT(DISTINCT(transactionId)) OVER (PARTITION BY date, clientId, DayTime) AS unique_transactions FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM `owox-analytics.t_kravchenko.Demo_data`)) GROUP BY date, clientId, DayTime, avg_revenue, max_revenue, min_revenue, sum_revenue, transactions, unique_transactions ORDER BY transactions DESC

    SELECT date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour FROM ( SELECT date, clientId, DayTime, hour, DENSE_RANK() OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS rank, revenue, LEAD( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lead_revenue, LAG( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lag_revenue, FIRST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS first_revenue_by_hour, LAST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS last_revenue_by_hour FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM [owox-analytics:t_kravchenko.Demo_data])) GROUP BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour ORDER BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour
SELECT date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour FROM ( SELECT date, clientId, DayTime, hour, DENSE_RANK() OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS rank, revenue, LEAD( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lead_revenue, LAG( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lag_revenue, FIRST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS first_revenue_by_hour, LAST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS last_revenue_by_hour FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM [owox-analytics:t_kravchenko.Demo_data])) GROUP BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour ORDER BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour

    SELECT date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour FROM ( SELECT date, clientId, DayTime, hour, DENSE_RANK() OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS rank, revenue, LEAD( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lead_revenue, LAG( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lag_revenue, FIRST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS first_revenue_by_hour, LAST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS last_revenue_by_hour FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM `owox-analytics.t_kravchenko.Demo_data`)) GROUP BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour ORDER BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour
SELECT date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour FROM ( SELECT date, clientId, DayTime, hour, DENSE_RANK() OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS rank, revenue, LEAD( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lead_revenue, LAG( revenue, 1) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS lag_revenue, FIRST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS first_revenue_by_hour, LAST_VALUE(revenue) OVER (PARTITION BY date, clientId, DayTime ORDER BY hour) AS last_revenue_by_hour FROM ( SELECT date, date_UTC, clientId, transactionId, revenue, page, hour, CASE WHEN hour>=9 AND hour<=18 THEN 'working hours' ELSE 'non-working hours' END AS DayTime FROM `owox-analytics.t_kravchenko.Demo_data`)) GROUP BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour ORDER BY date, clientId, DayTime, hour, rank, revenue, lead_revenue, lag_revenue, first_revenue_by_hour, last_revenue_by_hour

우리의 클라이언트
자라다 22% 더 빠름

마케팅에서 가장 효과적인 것을 측정하여 더 빠르게 성장

마케팅 효율성 분석, 성장 영역 찾기, ROI 증가

데모 받기

결론

독자를 위한 보너스

SQL 쿼리 실습을 위한 데모 데이터

지금 다운로드