指标解释

如何计算复购率/回购率

计算方法一：复购的人 复购率 = 单位时间内购买次数大于1的人/所有购买的人 例如： 一段时间内，10个人中有3个人购买2次，这3个人中有一个人又购买了一次，累计复购人数为3人，则这段时间内的复购率为30%。

计算方法二：复购次数 复购率 = 单位时间内复购次数/所有购买的人 例如： 一段时间内，10个人中有3个人购买2次，这3个人中有一个人又购买了一次，累计复购次数为4次，则这段时间内的复购率为40%。

复购和回购的区别

复购是一个单位时间内的多次购买，回购是在下一个单位时间内仍然购买。 例如： 6月总共100人购买产品，其中有10人在6月购买了2次，5人在6月购买了3次。 按照方法一计算，6月的复购率为15% 按照方法二计算，6月的复购率为20% 若6月购买过产品的100个人中有10个在7月又购买了，则7月的回购率为10%。

何时关注复购？

90天内重复购买率达到1%~15%；说明你处于用户获取模式；把更多的精力和资源投入到新用户获取和转化； 90天内重复购买率达到15~30%；说明你处于混合模式；平衡用在新用户转化和老用户留存、复购上的精力和资源； 90天内重复购买率达到30%以上；说明你处于忠诚度模式；把更多的精力和资源投入到用户复购上；上述观点出自《精益数据分析》，跟本人无关。

SQL题

题目出自《七周成为数据分析师》。

建表导入数据

DROP TABLE IF EXISTS orderinfo;

CREATE TABLE orderinfo ( id int, userid int, ispaid varchar(3), price decimal(10, 2), paidtime datetime NULL );

DROP TABLE IF EXISTS userinfo;

CREATE TABLE userinfo ( userid int, sex varchar(3) NULL, birth date NULL );

# 设置允许从本地导入文件

SET GLOBAL local_infile=1;

load data local infile 'F:/order_info_utf.csv' into table test.orderinfo fields terminated by ',' optionally enclosed by "'" escaped by '' lines terminated by '\r\n';

load data local infile 'F:/user_info_utf.csv' into table test.userinfo fields terminated by ',' optionally enclosed by "'" escaped by '' lines terminated by '\r\n' (userid,sex,birth);

我这边选择使用load data语句将csv导入mysql，你也可以选择使用可视化界面导入，例如SQLyog,Navicat等。语句导入的执行效率较高。下图为使用Navicat导入的效果。

下图为使用load data命令导入的效果。

1. 统计不同月份已支付的订单数，下单人数

SELECT CONCAT(YEAR(paidtime), '/', MONTH(paidtime)) AS 下单年月

, COUNT(*) AS 支付订单数, COUNT(DISTINCT userid) AS 下单人数

FROM orderinfo

WHERE ispaid = '已支付'

GROUP BY 下单年月;

2. 统计不同月份的回购率和复购率(复购率按照下单人数来算)

SELECT 下单年月, COUNT(c) AS 下单人数

, COUNT(if(c > 1, 1, NULL)) AS 当月复购人数

, concat(round(COUNT(if(c > 1, 1, NULL)) / COUNT(c) * 100, 2), '%') AS '当月复购率'

FROM (

SELECT CONCAT(YEAR(paidtime), '/', MONTH(paidtime)) AS 下单年月

, COUNT(userid) AS c

FROM orderinfo

WHERE ispaid = '已支付'

GROUP BY 下单年月, userid

) a

GROUP BY 下单年月;

SELECT a.date, COUNT(a.userid) AS 当月购买人数, COUNT(b.userid) AS 次月回购人数

, concat(round(COUNT(b.userid) / COUNT(a.userid) * 100, 2), '%') AS 次月回购率

FROM (

SELECT DATE_FORMAT(paidtime, '%Y-%m-%01') AS date, userid

FROM orderinfo

WHERE ispaid = '已支付'

GROUP BY date, userid

ORDER BY userid

) a

LEFT JOIN (

SELECT DATE_FORMAT(paidtime, '%Y-%m-%01') AS date, userid

FROM orderinfo

WHERE ispaid = '已支付'

GROUP BY date, userid

ORDER BY userid

) b

ON a.userid = b.userid

AND a.date = date_sub(b.date, INTERVAL 1 MONTH)

GROUP BY date

3. 统计多次消费的用户，第一次和最后一次消费时间的间隔

SELECT userid

, DATEDIFF(MAX(paidtime), MIN(paidtime)) AS 间隔天数

FROM orderinfo

WHERE ispaid = '已支付'

GROUP BY userid

HAVING COUNT(userid) > 1;

4. 统计男女消费频次是否有差异

SELECT sex, AVG(c) AS 消费频次

FROM (

SELECT COUNT(*) AS c, sex

FROM orderinfo o

INNER JOIN userinfo u ON o.userid = u.userid

WHERE sex IS NOT NULL

AND ispaid = '已支付'

GROUP BY o.userid

) a

GROUP BY sex

5. 统计男女消费金额是否有差异

SELECT sex, AVG(s) AS 消费金额

FROM (

SELECT SUM(price) AS s, sex

FROM orderinfo o

INNER JOIN userinfo u ON o.userid = u.userid

WHERE sex IS NOT NULL

AND ispaid = '已支付'

GROUP BY o.userid

) a

GROUP BY sex

6. 统计不同年龄段的用户消费金额是否有差异

SELECT CASE

WHEN age BETWEEN 0 AND 9 THEN '0-9岁'

WHEN age BETWEEN 10 AND 19 THEN '10-19岁'

WHEN age BETWEEN 20 AND 29 THEN '20-29岁'

WHEN age BETWEEN 30 AND 39 THEN '30-39岁'

WHEN age BETWEEN 40 AND 49 THEN '40-49岁'

WHEN age BETWEEN 50 AND 59 THEN '50-59岁'

WHEN age BETWEEN 60 AND 69 THEN '60-69岁'

WHEN age BETWEEN 70 AND 79 THEN '70-79岁'

WHEN age BETWEEN 80 AND 89 THEN '80-89岁'

ELSE NULL

END AS 年龄段, round(AVG(price), 2) AS 平均消费

FROM orderinfo o

INNER JOIN (

SELECT userid, year(now()) - year(birth) AS age

FROM userinfo

WHERE year(birth) > 1900

) a

ON (o.userid = a.userid

AND age IS NOT NULL

AND ispaid = '已支付')

GROUP BY 年龄段

HAVING 年龄段 IS NOT NULL

ORDER BY 年龄段

7. 统计消费的二八法则，消费的top20%用户，贡献了多少额度

SELECT @sum_price := SUM(price)

, @count_user := COUNT(DISTINCT userid)

FROM orderinfo

WHERE ispaid = '已支付';

SELECT SUM(s) AS '前20%累计消费'

, concat(round(SUM(s) * 100 / @sum_price, 2), '%') AS '占比'

FROM (

SELECT userid, SUM(price) AS s, row_number() OVER (ORDER BY SUM(price) DESC) AS r

FROM orderinfo

WHERE ispaid = '已支付'

GROUP BY userid

) a

WHERE a.r <= @count_user * 0.2

数据数据集及代码下载链接：百度云​pan.baidu.com

提取码：xarx