Without a pivot, you can use a cross join instead
Note this only works if you know how many names you will have everytime you run it and if each name only appears once in the original table.(otherwise the max function below is not appropriate)
create table #test(ID int, Name char(5), planned int, actual int, difference_between int)
insert into dbo.#test
values
(54, 'Name1', 0, 8975, -8957),
(54, 'Name2', 0, 5401, -5401),
(54, 'Name3', 0, NULL, NULL),
(54, 'Name4', 0, NULL, NULL)
select case t.occurno when 1 then 'Planned' when 2 then 'Actual' when 3 then 'Difference' end as [Key]
, max(case when Name = 'name1' then case t.occurno when 1 then planned when 2 then actual when 3 then difference_between else 0 end end) as Name1
, max(case when Name = 'name2' then case t.occurno when 1 then planned when 2 then actual when 3 then difference_between else 0 end end) as name2
, max(case when Name = 'name3' then case t.occurno when 1 then planned when 2 then actual when 3 then difference_between else 0 end end) as name3
, max(case when Name = 'name4' then case t.occurno when 1 then planned when 2 then actual when 3 then difference_between else 0 end end) as name4
from dbo.#test
cross join
(select top 3 ROW_NUMBER() over(order by occurno) as occurno
from (select 1 as occurno) t
group by cube(occurno,occurno,occurno,occurno)
) t
group by t.occurno
pivot.unpivotfirst to generate 12 rows withNameandStatuscolumns and all 12 values in distinct rows, and thenpivotbased off of that.