MySQL 8.0

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

CREATE TABLE t
    (ID_sale int, sales_person int, sale_date date)
;
    
INSERT INTO t
    (ID_sale, sales_person, sale_date)
VALUES
    (1, 50, '2022-10-19'),
    (2, 43, '2022-9-17'),
    (3, 50, '2022-3-15'),
    (4, 43, '2022-2-13'),
    (5, 50, '2022-1-22'),
    (6, 10, '2022-2-5'),
    (7, 12, '2022-1-7')
;

Records: 7  Duplicates: 0  Warnings: 0


select *
from   t;


ID_sale
sales_person
sale_date


1
50
2022-10-19

2
43
2022-09-17

3
50
2022-03-15

4
43
2022-02-13

5
50
2022-01-22

6
10
2022-02-05

7
12
2022-01-07


select    max(recent_ID_sale)                         as ID_sale
         ,sales_person
         ,max(recent_sale)                            as recent_sale
         ,coalesce(max(last_sale), max(recent_sale))  as last_sale
from
(
select    sales_person
         ,case when rank() over(partition by sales_person order by sale_date desc) = 1 then sale_date end as recent_sale
         ,case when rank() over(partition by sales_person order by sale_date desc) = 2 then sale_date end as last_sale
         ,case when rank() over(partition by sales_person order by sale_date desc) = 1 then ID_sale   end as recent_ID_sale

from      t
) t
group by sales_person
order by recent_sale desc


ID_sale
sales_person
recent_sale
last_sale


1
50
2022-10-19
2022-03-15

2
43
2022-09-17
2022-02-13

6
10
2022-02-05
2022-02-05

7
12
2022-01-07
2022-01-07

ID_sale	sales_person	sale_date
1	50	2022-10-19
2	43	2022-09-17
3	50	2022-03-15
4	43	2022-02-13
5	50	2022-01-22
6	10	2022-02-05
7	12	2022-01-07